burn-down Part A: 24 rows closed as fixed by later work, 2 duplicates merged, R-376/R-817/R-818 done (335 -> 306); Part A table
gates / gates (push) Failing after 1m16s
gates / gates (push) Failing after 1m16s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0159rPz1ZhFKsS53msqPYxtS
This commit is contained in:
1 parent
58ce696e16
commit
f5a0aeb0b8
8 files changed
+714
-38
No files matched your search
@@ -1,5 +1,14 @@
|
||||
# 08 — The app-down alarm ladder
|
||||
|
||||
> **How to read this document.** Where a statement is marked, it is marked like this — the same wording as
|
||||
> `07-backup-architecture.md:11-17`, carried here on 2026-10-05 (R-376, the three documents written after the
|
||||
> 2026-08-22 pass):
|
||||
>
|
||||
> - **[DESIGN]** — a decision taken. Not derived from code; the code may not implement it yet.
|
||||
> - **[FACT]** — an observed property, carrying a `file:line`, a live command output or a citation.
|
||||
>
|
||||
> **An unmarked statement means "not yet classified", never "observed"** (R-376).
|
||||
|
||||
**Written 2026-08-23, with controller v0.222.0 (R-384).**
|
||||
|
||||
**The absence is the finding.** Until this file existed, no document owned the question *"when does a
|
||||
|
||||
@@ -1,5 +1,14 @@
|
||||
# 09 — How an app update works, and what it is becoming
|
||||
|
||||
> **How to read this document.** Where a statement is marked, it is marked like this — the same wording as
|
||||
> `07-backup-architecture.md:11-17`, carried here on 2026-10-05 (R-376, the three documents written after the
|
||||
> 2026-08-22 pass):
|
||||
>
|
||||
> - **[DESIGN]** — a decision taken. Not derived from code; the code may not implement it yet.
|
||||
> - **[FACT]** — an observed property, carrying a `file:line`, a live command output or a citation.
|
||||
>
|
||||
> **An unmarked statement means "not yet classified", never "observed"** (R-376).
|
||||
|
||||
> **LIVING DOCUMENT. Every slice of the update arc updates this file in the same session.**
|
||||
> Opened 2026-09-02 with slices 1 and 2. Its absence was **R-438**: the update mechanism was chosen
|
||||
> deliberately and written down nowhere, which is how a deliberate design gets "fixed" by someone who
|
||||
@@ -620,6 +629,11 @@ R-636's louder repeated alarm.
|
||||
controller images are deleted by the same in-use rule as decision 53 — *operator ruling 2026-10-01 (R-745, option 3A).*
|
||||
**Why:** about 50 old controller versions sat on each demo box (R-745); a release is ~400 MB unpacked and several ship
|
||||
a day. Registry tags are never deleted by this.
|
||||
*Clarified 2026-10-05 (R-817), the ruling unchanged:* a swap records the image RUNNING when it starts
|
||||
(`felhom-agent internal/localapi/controllerswap.go:236-240`, `st.Previous`) and a failed swap writes exactly that
|
||||
image back (`:289`). After a good swap that image is "the one before" the running one — so „the one before it" here
|
||||
and R-745's „rolls back to the RUNNING image" name the same image, seen before and after the swap. The controller
|
||||
never hands the agent an older roll-back target.
|
||||
|
||||
### 2026-10-01 — decided by CC unattended, operator may reverse
|
||||
|
||||
|
||||
@@ -1,5 +1,14 @@
|
||||
# 11 — Operating-system updates: the host, the guest and the Docker engine
|
||||
|
||||
> **How to read this document.** Where a statement is marked, it is marked like this — the same wording as
|
||||
> `07-backup-architecture.md:11-17`, carried here on 2026-10-05 (R-376, the three documents written after the
|
||||
> 2026-08-22 pass):
|
||||
>
|
||||
> - **[DESIGN]** — a decision taken. Not derived from code; the code may not implement it yet.
|
||||
> - **[FACT]** — an observed property, carrying a `file:line`, a live command output or a citation.
|
||||
>
|
||||
> **An unmarked statement means "not yet classified", never "observed"** (R-376).
|
||||
|
||||
> | | |
|
||||
> |---|---|
|
||||
> | **Status** | **NOT RATIFIED — a PROPOSAL with operator rulings, corrected by the 2026-10-04 spike (§7.1, C1–C12); §8 step 2 BUILT 2026-10-04 (§8.1).** Ratification is Viktor's review, not an editor's. |
|
||||
|
||||
@@ -0,0 +1,317 @@
|
||||
{"id": "R-10", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/appbackup/dbdump.go:364 `if err := tmpFile.Sync(); err != nil {` then :390 `if err := os.Rename(tmpPath, finalPath); err != nil {` with no directory Sync after; the twin at controller/internal/backup/backup.go:948 `_ = dir.Sync()` does sync the dir. Origin: audits/CAMPAIGN-6E-2026-07-15.md:129.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/appbackup/dbdump.go"], "change": "After the os.Rename in DumpOne, open filepath.Dir(finalPath) and call a best-effort dir.Sync() (log at DEBUG on error), mirroring atomicPromoteTar in backup.go:948.", "test": "Unit test that DumpOne still produces the final file and leaves no .tmp; fsync itself is not observable in a unit test, so add a small syncDir seam and assert it is called with the dump directory (red-proof by removing the call).", "minutes": 30}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-25", "sev": "P4", "category": "Storage & devices", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/web/storage_handlers.go:153 `uuid := resolveEnrollUUID(ctx, agent, device)` still resolves by device PATH after format, then AssignDisk(uuid) at the next step; FormatResult (controller/internal/agentapi/client.go:384-395) carries DurableID only for the confirmation path, not the new fs UUID. Binding resolve+assign to the format's durable-id needs the agent to return the new fs identity (two repos).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-76", "sev": "P4", "category": "Apps & catalog", "group": "UNCHECKED", "evidence": "Behaviour is FileBrowser-image runtime behaviour (mode/setgid of UI-created folders), only observable on a live box. The image has changed since the finding: controller/internal/infra/infra.go:27 `FileBrowserImage = \"gtstef/filebrowser:1.5.6-stable\"` (finding was on 1.3.3). The comment at infra.go:207-208 still asserts `umask 002 so folders the customer creates here come out group-writable (2775 with the parent's setgid)`, which the row says was false on 1.3.3 — needs a re-measure on 1.5.6 before deciding. Note: the register row itself is truncated mid-sentence (\"does not say t\").", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-89", "sev": "P4", "category": "Business & legal", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No retention policy object in hub: `grep -rln -i 'retentionpolicy|retention_policy' felhom.eu/hub` returns nothing (felhom.eu@53d8131b). Commercial per-customer policy = money/product decision + new reconciler.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-91", "sev": "P4", "category": "Backup & restore", "group": "UNCHECKED", "evidence": "Whether /srv/pbs-felhom still exists on ep0 is live-only (ep0 is protected; not touched). Source-side: CONTEXT.md:3656 still reads \"`/srv/pbs-felhom` is 13 G of dead weight on `/` awaiting R-91's go-ahead\". Extra fact found: documentation/runbooks/offsite-endpoint.md:24 still says the datastore `felhom-offsite` is at `/srv/pbs-felhom` and :119 `proxmox-backup-manager datastore create felhom-offsite /srv/pbs-felhom`, contradicting RUNBOOK-ep0-datastore-volume-2026-07-27.md:8 (moved to /mnt/pbs-datastore). The row's CONTEXT.md:1018 citation is stale (now :3656).", "dup_of": null, "unique_facts": "offsite-endpoint.md:24 and :119 still name /srv/pbs-felhom as the live datastore path (stale since the 2026-07-27 move to /mnt/pbs-datastore) — a doc fix independent of the deletion; CONTEXT citation moved from :1018 to :3656.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-92", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu@53d8131b hub/internal/web/pbsdr_box.go:57 and :64 `view.UsedStr = fmtBytesGB(snap.UsedBytes)`; hub/internal/web/offsite_box.go:54 `return fmt.Sprintf(\"%.1f GB\", float64(b)/float64(int64(1)<<30))` — still 0.1 GB granular. Note the row's own trigger (\"when retention becomes customer-visible\") has not fired.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/web/pbsdr_box.go", "hub/internal/web/offsite_box.go", "hub/internal/web/templates/offsite.html"], "change": "Add an exact-bytes value to the PBS DR view (e.g. UsedBytesExact rendered as a title= tooltip or a MB-precision string below 10 GB) without changing fmtBytesGB for other callers.", "test": "Table test on the view builder: two snapshots 50 MB apart render different exact strings; render test of offsite.html shows the exact value.", "minutes": 30}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-93", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Premise gone per the row itself (R-461, CLOSED-ITEMS.md:636): drill-r50 VM no longer exists on either demo box. target-selection.md:111 keeps the fence with the note \"the VM does not exist anywhere, so the fence currently protects nothing\". Only remaining references are comments/tests (hub/internal/monitor/deadline_anchor_test.go:16, deadline_tiers.go:59).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A row about choosing between two fixtures, neither of which exists any more.", "cost": "Building a new synthetic drift fixture is a design task (M), not a fix of this row.", "if_never": "Nothing breaks; there is no drift fixture either way. If one is wanted, it is a new row.", "pick": "close-as-accepted (operator word needed: close, or reopen as 'build a drift fixture'); also drop the dead drill-r50 fence in target-selection.md:111 at close"}, "minutes_spent": 4}
|
||||
{"id": "R-99", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No phantom-snapshot cleanup in felhom.eu/hub or felhom-agent (grep -i phantom finds only agent runner/test detection code; no removal path). Deletion on a customer datastore is a separate operator ruling per the row — customer data.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-104", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/backup/offbox.go:193-222 ClassifyOffsiteFailure has cases NoUnits/NoRepo/Transport and `default: return OffsiteFailUnknown` — no lock case, although offbox.go:826 already defines `var offboxLockRe = regexp.MustCompile(`repository is already locked`)`. The self-heal half is built (offbox.go:839-858 unlock --remove-all + retry once), as the row's 2026-08-22 note says.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/backup/offbox.go", "controller/internal/i18n/locales/hu.json", "controller/internal/i18n/locales/en.json", "controller/internal/backup/*_test.go"], "change": "Add an OffsiteFailLocked class matched by offboxLockRe in ClassifyOffsiteFailure (before transport) and a cause line in OffsiteFailureMessage telling the operator the repository is locked by an interrupted run and how it clears.", "test": "Table test: a restic 'repository is already locked' error classifies as Locked (red-proof: fails today as Unknown); i18n parity gate for the new key.", "minutes": 50}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-124", "sev": "P4", "category": "Backup & restore", "group": "NOT-WORTH-IT", "evidence": "Still true: felhom-agent@e06ed97 internal/hub/dr_recipe.go:61 `const PBSRootNamespace = \"root\"` and :276 `c.Namespace = PBSRootNamespace`; agent internal/pbs/report.go:24 `ns = \"root\"`. The comment at dr_recipe.go:57-58 already documents that PBS spells it \"\".", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The disaster-recovery recipe writes the PBS root namespace as the word 'root', but PBS itself uses an empty name, so a pasted '--ns root' fails.", "cost": "Changing it alters a wire field read by the hub (cross-repo wire contract + recipe producers), for a case no customer has: every box writes a per-customer namespace.", "if_never": "An operator restoring a box with NO namespace line would get one failed command and have to drop --ns; no data risk. The constant's comment already warns.", "pick": "close-as-accepted"}, "minutes_spent": 4}
|
||||
{"id": "R-129", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "Docs still say no key: felhom.eu@53d8131b documentation/operations/nodes.md:110 `### Access — there is no baked SSH key` and :112 \"no operator public key is on this box\"; target-selection.md:111 still flags R-129 unresolved; MEMORY.md:34 says `ssh demo-hp`, NO KEY→G1. Whether the key works today is a live fact (not checked — no ssh in this pass).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["documentation/operations/nodes.md", "documentation/runbooks/target-selection.md"], "change": "After one read-only `ssh -o BatchMode=yes demo-hp true` (and reading root's authorized_keys comment to name the key), rewrite nodes.md 'Access' section to the measured truth and drop the R-129 caveat in target-selection.md:111-112 (also update the memory index line).", "test": "Positive control: the BatchMode ssh succeeds/fails as the doc now states; repo_gates.py doc gates pass.", "minutes": 30}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-134", "sev": "P4", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu@53d8131b hub/internal/cloudflare/unblock.go:117 `for _, name := range []string{domain, parentDomain(domain)} {` and :136-141 parentDomain strips exactly one label (`strings.SplitN(domain, \".\", 2)`); controller strips progressively (controller/internal/cloudflare/zone.go per row).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/cloudflare/unblock.go", "hub/internal/cloudflare/unblock_test.go (new)"], "change": "Extract a pure zoneCandidates(domain) []string that yields the name and every parent down to two labels, and loop resolveZone over it (same order: most specific first).", "test": "Table test on zoneCandidates: 'a.b.felhom.eu' yields [a.b.felhom.eu b.felhom.eu felhom.eu] (red-proof: one-label version yields only two); no HTTP needed because apiBase is a const.", "minutes": 40}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-161", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Automatic half exists: app-catalog-felhom.eu@917a779 .gitea/workflows/gates.yml:40 `run: cd ws/app-catalog-felhom.eu && python3 scripts/catalog_gates.py --fast`; .githooks/pre-push:84 runs the same. Only the runtime volume-persistence gate stays a manual periodic run, by operator ruling 2026-08-02.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The runtime check that app data lands on a volume is run by hand, not on every push.", "cost": "Automating it means CI pulling and starting ~53+ app images per push — slow, and the row itself says such CI gets disabled.", "if_never": "A template that writes data outside its volume can ship until the next periodic run catches it; the static gates and pre-push still run.", "pick": "close-as-accepted (residual is a deliberate ruling; owner operator)"}, "minutes_spent": 3}
|
||||
{"id": "R-162", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Gate is app-catalog-felhom.eu scripts/check-volume-persistence.py (`docker diff` at :41, :72); behaviour on a non-overlay driver is not reachable from source and the row says it fails closed. Status WATCHING, no defect.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "If Docker ever ran on a storage driver where `docker diff` does not work, the persistence gate would refuse to report and blame the prober instead of the driver.", "cost": "A driver probe + reworded message in the catalog script, ~1 h, for a driver nobody runs.", "if_never": "Nothing, unless a non-overlay driver ships; even then the gate fails closed (no false green).", "pick": "close-as-accepted"}, "minutes_spent": 3}
|
||||
{"id": "R-164", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Predicate still absent: felhom-controller controller/internal/appbackup/dbdump.go:544 still only WARNs `its accounts table has NO rows`; restore still replays dump + tar (internal/backup/restore_unit.go:114-118 hasReplayableDump). Blocked on a design (live-vs-dump per-table counts).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-169", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Working-style ruling owed by operator; nothing in source to fix. Current nets per row: pre-push hooks + Gitea runner alarm (e.g. app-catalog .gitea/workflows/gates.yml:40).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "CI only reports after a push lands, because every repo pushes straight to main with no pull request.", "cost": "Making CI blocking needs branch protection plus a PR workflow for every change — a slower way of working for a one-operator project.", "if_never": "A `--no-verify` push can land broken code until the operator reads the CI alarm e-mail.", "pick": "close-as-accepted (row itself says decide only if the window ever costs something)"}, "minutes_spent": 2}
|
||||
{"id": "R-177", "sev": "P4", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller@7690c27 controller/cmd/controller/main.go:1546 `sched.Daily(\"fill-watch\", \"03:30\", func(ctx context.Context) error { return fillWatcher.Check() })`; internal/scheduler/scheduler.go:269 has GetJobs but grep finds no RunNow/Trigger method and no run-job route in internal/web. Needs a new operator-gated trigger endpoint (auth surface) — a new mechanism, solve together with R-279.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-184", "sev": "P4", "category": "Box system & updates", "group": "FIXED-BY-LATER-WORK", "evidence": "Fixed by felhom.eu b55fc17d \"hub v0.102.0 — refuse to vouch a version that cannot be installed (R-273)\" — exactly shape (b), validate at vouch time in the hub. felhom.eu/hub/internal/web/configs.go:1358 `res := s.gitea.PackageDownloadable(ctx, t.pkg, t.version, t.file)` and :1365 `s.logger.Printf(\"[WARN] artifact vouch REFUSED: %s package %s is NOT downloadable (R-287)\", ...)`; tag leg at :1343 TagServesFile; unreachable registry also refuses.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-194", "sev": "P4", "category": "Box system & updates", "group": "NOT-WORTH-IT", "evidence": "PVE behaviour, not our code; row states the self-repair already tolerates it (fires on the next probe after the cache clears). No source change to check.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Proxmox caches permissions, so a removed storage grant can still read as present for seconds to minutes; our self-repair notices only after the cache expires.", "cost": "Adding a second signal (storage content listing) to the agent's grant probe is a new mechanism, needs live measurement on a box.", "if_never": "A lost grant is noticed up to ~16 min late; the repair still happens on its own.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-206", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "homelab-manifests@87dfc29 (/home/kisfenyo/git/homelab-manifests): no daemon.json template in homelab-ansible (grep finds only a comment at roles/node_housekeeping/templates/node-housekeeping.sh.j2:17 and homelab-ansible/CLAUDE.md:54). Part (b) was superseded by fc9fbb8 (\"correct the expired Docker rationale\"): the script now says at :13-20 do NOT add docker calls, the GC policy in daemon.json is the control point. DooPlex work — not unprompted.", "dup_of": null, "unique_facts": "Part (b) (prune in the role) is superseded by fc9fbb8 — the role now deliberately forbids docker calls and names daemon.json GC policy as the control; remaining scope is (a) template daemon.json in Ansible + (c) restart-and-verify.", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-207", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "Fixed by homelab-manifests fc9fbb8 \"node_housekeeping: guard DRY_RUN, correct the expired Docker rationale, pin container log rotation\". /home/kisfenyo/git/homelab-manifests/homelab-ansible/roles/node_housekeeping/templates/node-housekeeping.sh.j2:137 `if [[ \"${DRY_RUN}\" == \"1\" ]]; then` inside write_metrics, :138 logs \"file left untouched\".", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-208", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/Dockerfile:12 `ARG VERSION=dev` and :13 `ARG GIT_COMMIT=unknown` sit above :19 `RUN go mod download || true`; felhom.eu@53d8131b hub/Dockerfile:3 `ARG VERSION=dev`, :4 `ARG BUILD_TIME=unknown` above :9 `RUN go mod download || true`.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller (and the identical one-line move in felhom.eu/hub — two repos, each trivial)", "files": ["controller/Dockerfile", "felhom.eu: hub/Dockerfile"], "change": "Move the ARG VERSION/GIT_COMMIT (controller) and ARG VERSION/BUILD_TIME (hub) declarations down to just above the final `go build` RUN.", "test": "Build twice with different --build-arg VERSION on a clean tree; the second build must show `RUN go mod download` CACHED; `--version` of the built binary still shows the passed version.", "minutes": 30}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-209a", "sev": "P4", "category": "Process & tooling", "group": "UNCHECKED", "evidence": "Live-only: whether DooPlex has rebooted and /var/log/felhom-store-postboot-check.log says PASS. Not read (DooPlex is Tier 2, operator ruled no reboot; this pass touches no machine). No source claim to check.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-210", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Operator ruling owed; row records CC's view 'not worth doing for the space' (~27 GB reclaimable vs 199 GB free). Workspace CLAUDE.md also forbids `docker image prune -a` on DooPlex.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "193 old controller/hub images exist only on DooPlex and cannot be re-pulled; the question is whether to delete them.", "cost": "An operator decision plus a careful targeted delete on the production host; returns ~27 GB.", "if_never": "~27 GB stays used on a disk with ~199 GB free; old images remain as clutter (and as the only copies of very old builds).", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-213", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Row is a not-started design (live-vs-backup comparison, then put-back flow), operator-owned; nothing in source to verify against.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-230", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Owed rulings, not code: (a) bulk-correction ruling on MEMORY.md staleness (MEMORY.md index still carries version literals, e.g. 'ctrl 0.224.0', 'hub 0.109.0'); (c) spec-as-failing-test pilot not started. (b) closed. Operator decision required.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-246", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu@53d8131b hub/internal/store/store.go:3248 `func (s *Store) MarkEscrowStale(hostID string) error {` still has no production caller (grep: only definition + comments at offsite.go:208,216); stale_at still read (store.go:3182 clears it). Ruling owed by operator: evidential setter or retire the column (folds R-248).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-256", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/i18n/locales/hu.json:1406 `\"flash.offbox.mgr_unavailable\": \"A mentéskezelő nem elérhető.\",` used at controller/internal/web/offbox_handlers.go:54 and :197; sibling :1407 mgr_unreachable used at offbox_handlers.go:582; en.json:1415-1416 same shape.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/i18n/locales/hu.json", "controller/internal/i18n/locales/en.json"], "change": "Rewrite flash.offbox.mgr_unavailable / mgr_unreachable in both languages to say the backup service is not running yet and give a route (try again in a few minutes; if it persists, contact support).", "test": "i18n parity/accent gates (controller_gates.py) pass; a handler test with nil backupMgr asserts the redirect carries the key (exists or add one). Owner is operator (copy) — needs a nod on the wording.", "minutes": 20}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-261", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu@53d8131b hub/internal/store/selfbind.go:111 `func (s *Store) CountSelfBindTokens(customerID string) (int, error) {`; only callers are hub/internal/web/customer_delete_test.go:511 and selfbind_automint_test.go:29 — no production caller.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/store/selfbind.go"], "change": "Reword the doc comment (selfbind.go:106-110) to say it is a test accessor and name the two tests that pin the auto-mint invariant (selfbind_automint_test.go, customer_delete_test.go) — or, if the operator prefers, add one post-mint production check that logs [WARN] when count != 1.", "test": "Comment-only option: existing tests stay green; production-check option: unit test that a pre-seeded extra token produces the WARN (red-proof).", "minutes": 20}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-263", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/settings/settings.go:1655 `// from every other. This is the ONLY writer of StoragePath.BackupTarget — registration must never set` while :1699 `s.StoragePaths[i].BackupTarget = false` (ClearBackupTarget) also writes it; :1684 is SetBackupTarget's write.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/settings/settings.go", "controller/internal/settings/backup_target_role_test.go"], "change": "Change the comment to 'the only writer that GRANTS the role' and add a source-scanning test that finds every `.BackupTarget =` assignment in non-test settings code and fails if any other than SetBackupTarget can assign a non-false value.", "test": "The new test passes today; red-proof by planting a temporary `BackupTarget = true` in another function and seeing it fail.", "minutes": 45}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-264", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu@53d8131b scripts/wire_contract_gate.py still allowlists the six with _R264: :242 selfupdate_pending, :246 selfupdate_pending_version, :255 restore_tests.mount_parity, :258 restore_tests.mount_inventory, :281 backup.last_db_dump, :282 backup.last_integrity_check. Each reader is a design per the row.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-266", "sev": "P4", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/report/builder.go:94 `{Mount: \"/\", Label: \"SSD\", TotalGB: sysInfo.DiskTotalGB, UsedGB: sysInfo.DiskUsedGB, Percent: sysInfo.DiskPercent},` — no disk_known on the storage entry; hub has no disk_known (grep empty). Two-repo wire change gated by wire_contract_gate.py.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-279", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No operator/hub path to start an off-site run: grep for offbox run triggers in felhom.eu/hub/internal finds nothing; the only run entry is the customer dashboard handler (felhom-controller controller/internal/web/offbox_handlers.go:270 `if !s.backupMgr.OffboxRunnable() {`). Needs a new operator-authenticated trigger — sibling of R-177, not a duplicate (different job).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-284", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "Not a defect: felhom-controller@7690c27 controller/internal/web/templates/deploy.html:622 `<div id=\"storage-space-warn\" class=\"form-hint\" style=\"color:var(--warn);display:none\">` (hidden by default) and :808 `warn.style.display = freePct < 20 ? 'block' : 'none';` — correct direction. `git log -S \"freePct < 20\"` shows it unchanged since 69698a8 (v0.10.0), and deploy.html at c732fe1 (main as of 2026-08-09) already had display:none at :589. The 2026-08-09 report read raw HTML without running scripts.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A reported 'almost full' warning on an empty disk; the code shows the warning only below 20% free and hides it by default, so the report was a reading of unrendered HTML.", "cost": "Nothing to fix; a JS render test would need a browser harness the project does not have.", "if_never": "Nothing — the warning never showed on a 93%-free disk.", "pick": "close-as-accepted (close as not-a-defect)"}, "minutes_spent": 5}
|
||||
{"id": "R-285", "sev": "P4", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No maintenance/expected-downtime concept in hub: `grep -rln -i 'maintenance|expected_downtime|quiet_until|snooze' felhom.eu/hub/internal` returns nothing. New mechanism (M).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-286", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "Lesson (a) not written anywhere: grep -i 'different channel|same channel|independent channel' over documentation/runbooks/workspace-CLAUDE.md, felhom.eu/skills/*, .claude/rules/* returns nothing; felhom.eu/skills/felhom-evidence/SKILL.md:52 has the positive-control rule ('Plant the thing, find it...') but not the different-channel requirement. Part (b): RUNBOOK-hub-db-offsite-backup.md:101 copies a snapshot ($SNAP) not bare hub.db, so no bare-`cat hub.db` runbook found in runbooks/.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["skills/felhom-evidence/SKILL.md (or documentation/runbooks/workspace-CLAUDE.md standing rule 3)"], "change": "Add one paragraph: a positive control must come from a different channel than the measurement (different query path, snapshot, API or clock); give the 2026-08-09 stale-snapshot case as the example. Put it in ONE home (pointer elsewhere).", "test": "python3 scripts/check_skills.py and repo_gates.py pass; re-read the skill to confirm it loads.", "minutes": 25}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-287", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "Deleter established 2026-08-10 (R-267 newest-10 prune, recorded in the row itself); CI fixed by felhom-agent 53d047a \"Two guards, one number: bound the published check to the retention it must live with\" (R-291). felhom-agent@e06ed97 scripts/check-published-versions.py:101 `RETENTION_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), \"retention-policy.json\")`, :213 `keep = retention_kept()`; scripts/retention-policy.json:37 `\"generic_versions_kept\": 10,`. Follow-up row R-291 is open (OPEN-ITEMS.md:439).", "dup_of": null, "unique_facts": "scripts/retention-policy.json _comment lines ~15-17 still say no register row records a package prune and 'Container packages currently hold 19 each' — both withdrawn by R-287 (prune is in R-267; 19 was an unpaginated count, real 270/169). Move this stale-comment fix into R-291.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-288", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu@53d8131b documentation/architecture/00-capability-map.md is now 210 125 bytes / 30 937 words / 253 lines (`wc`), larger than the 134 642 bytes measured in the row; :38 still reads `*Verified 2026-07-16 against evidence corpus @ felhom.eu tip `4b18cc5``. Restructure is an M doc surgery, operator-owned.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-289", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "R-182 was closed by felhom.eu ef6ac6fe (2026-08-22, register compression): documentation/backlog/CLOSED-ITEMS.md:474 `| **R-182** | ... | **CLOSED — SHIPPED** (controller v0.194.0 + hub v0.90.0/.1, 2026-08-03) |`. The residue (digest never seen delivering) was since observed: documentation/audits/DRILL-chaos-night-2026-09-17.md:181 `backup_run_failures` „1 of 12 apps failed to back up in this nightly run: nextcloud\" listed as an alarm that fired and was true.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-290", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Gate exists (felhom.eu scripts/check_stands.py) but the map itself still carries the claims: documentation/architecture/00-capability-map.md has 95 'PROVEN-LIVE' occurrences (grep -c); demoting 12 rows or writing walk documents is M and blocked on R-288 per the row.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-291", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-agent/scripts/retention-policy.json still says the 10 is 'NOT a ruling anyone has been able to locate' and recorded_by: \"CC, from the registry's observed state; NOT from a located operator ruling\" — but R-287 (OPEN-ITEMS.md:424) records it IS the operator's newest-10 rule executed under R-267. The file also lists a reader 'documentation/runbooks/registry-retention.md (felhom.eu)' that does not exist (find under felhom.eu/documentation returns no such file). check-published-versions.py:101-104 reads the number. The deeper min_agent-floor bound still needs hub network (not small, recorded only).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-agent", "files": ["scripts/retention-policy.json"], "change": "Rewrite the _comment/recorded_by to cite the operator's newest-10 rule (R-267/R-287) instead of 'observed, not a ruling', and drop or correct the non-existent registry-retention.md reader. Keep the min_agent-floor note as the recorded better bound; then close R-291.", "test": "python3 scripts/agent_gates.py --fast (check-published-versions reads the file; JSON must still parse and generic_versions_kept stay 10)", "minutes": 20}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-292", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "hub/internal/web/templates/configuration.html:55 still reads 'the Gitea sha lookup failed (version missing / Gitea unreachable) or the manually-entered sha is invalid'; configs.go:1375 and :1395 both redirect to flash=artifact_sha_invalid; resolveArtifactSHA returns only (sha, ok bool) so the cause is lost.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/internal/web/configs.go", "hub/internal/web/templates/configuration.html"], "change": "Make resolveArtifactSHA return a reason (not-found / unreachable / bad manual sha) and redirect to three distinct flashes (reuse artifact_unverifiable for unreachable, add artifact_version_missing, keep artifact_sha_invalid for a bad typed sha incl. the wrapper sha at :1395).", "test": "Handler test per cause asserting the redirect flash, with a fake gitea returning 404 / network error and a malformed manual sha; red-proof by running against old code.", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-310", "sev": "P4", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/felhom-host-install.sh:3060 sets GOLDEN_CHECK_WHY=\"it is controller $ver, but the vouched golden is $ART_GOLDEN_VER\" and :3080-3081 die \"...${GOLDEN_CHECK_WHY}.\\n The vouched golden is ${ART_GOLDEN_VER:-<unknown>}.\" — duplicate stands. :984 still reads the vmid confirm from /dev/tty; no runbook (day0-install.md, RUNBOOK-byo/appliance-deployment.md mention --uninstall) names the pty requirement.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh", "documentation/runbooks/day0-install.md"], "change": "Drop the second 'The vouched golden is' sentence when GOLDEN_CHECK_WHY already names it (or drop the version from :3060); add one runbook line: --uninstall needs an interactive terminal; --force does not bypass the typed vmid confirm.", "test": "bash -n on the script + python3 scripts/repo_gates.py --fast (hostinstall gate); grep the die text renders the version once.", "minutes": 20}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-315", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/scripts/wire_contract_gate.py:30-48 still documents the test as a repo-wide literal-tag search ('IT PROVES REACHABILITY OF A NAME'); ROOTS at :88 includes the R-311 escrow/retained root. No receiver-type field-by-field comparison exists. Fix requires resolving receiver mirror types — a new mechanism (M).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-325", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/scripts/retrieval_promise_gate.py:54 still has its own literal STEMS = [\"visszaállíthat\", ...]; felhom.eu/scripts/hub_copy_gate.py:193-205 still reads it as a drift check. controller_gates.py:48-53 already imports shared scripts from the felhom.eu sibling, so the pattern exists.", "dup_of": null, "unique_facts": "Two-repo sequencing: hub_copy_gate.py:203-205 returns 'drift' when it cannot find a STEMS list, so the felhom.eu drift check must be removed in step with (or tolerate) the controller change.", "small_fix": {"repo": "felhom-controller", "files": ["controller/scripts/retrieval_promise_gate.py"], "change": "Import RETRIEVAL_STEMS from ../felhom.eu/scripts/customer_copy_vocab.py (same sibling-path pattern as controller_gates.py:48) and delete the STEMS literal; absent sibling = INCONCLUSIVE exit 2. Follow-up (felhom.eu, separate commit): remove hub_copy_gate.py's drift check, which would then fail to find STEMS.", "test": "python3 controller/scripts/controller_gates.py --fast; red-proof: remove a stem from the shared list and see the controller gate's decoy convict.", "minutes": 40}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-327", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/documentation/architecture/where-felhom-stands.yaml:126-130 still: id claim.code-naming, title \"The same word is used for two different secrets across three surfaces; the email points at a page a rebuilt machine does not show\", status: partial. Needs the operator's capability-map ruling first (dataset may not be raised on its own).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-331", "sev": "P4", "category": "Storage & devices", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/internal/agentapi/diskverdict.go:34 uncorrectableFailCount = 64; :33 comment still defers 'growth-rate detection once the box keeps history'. No growth-rate rule found. New mechanism (M).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-336", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No source change reduces the ep0 poll rate (pvestatd interval is Proxmox-side, not in our repos). Design question (does the hub need a 15-min fill reading) remains; acceptance needs an ep0 access-log measurement. Scaling item, not small.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-337", "sev": "P4", "category": "Monitoring & notifications", "group": "UNCHECKED", "evidence": "Live-only behaviour (WATCHING). From source: felhom-agent/internal/localapi/server.go:518 serves GET /backup/status and :1258 answers from s.pickLatestBackup (the in-memory store), which suggests collection cadence, but the refresh path after an out-of-schedule run was not established within the time box.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-345", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/hub/Makefile:21 'docker tag $(IMAGE):$(VERSION) $(IMAGE):latest' and :22 'docker push $(IMAGE):latest' still present; only commit touching the Makefile is 77b5a4ce (initial).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/Makefile"], "change": "Delete lines 21-22 (or move them behind an explicitly named opt-in target with a comment). Whether a stale :latest already sits on the registry is a separate live check for a session allowed to query it.", "test": "make -n docker-push | grep -c latest == 0", "minutes": 10}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-346", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "grep for ActiveEnterTimestamp|ExecMainStartTimestamp|InactiveExitTimestamp across felhom.eu/scripts, hub, felhom-agent, felhom-controller/controller, homelab-manifests (*.sh/*.go/*.py) returns ZERO hits; the R-341 row (now in CLOSED-ITEMS.md:211) carries the lstart reasoning. The row's only remaining action (check for other systemd-timestamp anchors) is answered: none exist.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A warning that a future reader might anchor an uptime slope on systemd's ActiveEnterTimestamp instead of the process start time.", "cost": "Nothing left to fix; the audit the row asked for finds no such use anywhere.", "if_never": "A future script could make the mistake; there is no current instance.", "pick": "close-as-accepted (audit done, zero instances)"}, "minutes_spent": 4}
|
||||
{"id": "R-348", "sev": "P4", "category": "Monitoring & notifications", "group": "STILL-TRUE-SMALL", "evidence": "felhom-agent/internal/backup/store.go:28 still reads '// Backups are unaffected — their freshness has a ground truth on the storage (R-84).' The hub-side pin the row asks for ALREADY EXISTS: felhom.eu/hub/internal/monitor/deadline_anchor_test.go:52 TestBackupFreshness_AgentRestartBlindWindow_NoAlarm, :295 TestNewestBackupEvidence_ReachesPastEmptyReports, :366 TestCheckBackupDeadlines_RestartBlindWindow_NoEvent; constant at hub/internal/monitor/deadline.go:36 backupEvidenceLookback.", "dup_of": null, "unique_facts": "Hub pin test already exists (deadline_anchor_test.go:52/:295/:366) — only the agent comment remains.", "small_fix": {"repo": "felhom-agent", "files": ["internal/backup/store.go"], "change": "Reword the comment: the backups list IS lost on restart and refills only when a backup runs; the hub's freshness VERDICT is unaffected because it looks back 7 days (hub monitor/deadline.go backupEvidenceLookback, pinned by deadline_anchor_test.go TestCheckBackupDeadlines_RestartBlindWindow_NoEvent).", "test": "Comment-only; go vet ./internal/backup; the hub tests named above already pin the consequence.", "minutes": 10}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-352", "sev": "P4", "category": "Storage & devices", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Placement half is an open operator ruling (SPEC-app-data-placement-2026-08-21.md, 'Viktor rules'); deploy route still has no server-side default: GetDefaultStoragePath has no caller in internal/stacks (grep returns only internal/api/router.go:1137 systemInfo). Point (2) is carried by R-368.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-364", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "No helper exists: ls felhom.eu/scripts shows nothing grep/accent-related, and no script mentions '0x80' or 'negative control'. The proposal is a self-contained tool.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/hu_grep.py (new)", "scripts/test_hu_grep.py (new)"], "change": "A helper that, for a pattern containing a byte >= 0x80, also runs an ASCII anchor (must hit) and a negative control (must miss) and refuses to print a zero unless both behave; documented in the felhom-evidence or ui-hungarian rule as the way to search Hungarian text.", "test": "Unit test with a temp file: accented string present -> count; transformed (e.g. octal-escaped) input -> REFUSED not zero; red-proof by disabling the anchor check.", "minutes": 60}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-365", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/i18n/locales/hu.json:368 'A kérésed szerint a korábbi távoli mentéseidet <strong>{{.AbandonDate}}</strong> napján véglegesen töröljük (még ... nap)'; handlers.go:1164-1167 sets AbandonDate/DaysLeft with no overdue branch; backups_remote.html:189 renders it unconditionally.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/handlers.go", "controller/internal/web/templates/backups_remote.html", "controller/internal/i18n/locales/hu.json", "controller/internal/i18n/locales/en.json"], "change": "Set AbandonOverdue when DueAt is past and render a new key ('a törlés esedékes, a következő napi karbantartáskor lefut' / English twin) instead of the future-tense sentence.", "test": "Render test with a clock past DueAt asserting the overdue key and not 'töröljük'; i18n parity + copy gates via controller_gates.py --fast; search with ASCII fragments.", "minutes": 45}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-367", "sev": "P4", "category": "Backup & restore", "group": "NOT-WORTH-IT", "evidence": "Prune guard confirmed still present: felhom-controller/controller/internal/backup/backup.go:1438 'continue // GUARD: an undeployed app's last backup is still its restore point'. The stranded file itself is on demo-hp (Tier 0) and was not checked live.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "One old 312 KB paperless database dump on the demo-hp test box sits under the app's old folder name; nothing reads or deletes it.", "cost": "An operator ruling plus a live hand-move or delete on one demo box, and an explanation that the adopted dump would not match the volume copies.", "if_never": "A small file stays on a disposable demo box; no customer is affected and the fix for new dumps already shipped (R-355).", "pick": "close-as-accepted (or delete it the next time the box is reprovisioned)"}, "minutes_spent": 4}
|
||||
{"id": "R-368", "sev": "P4", "category": "Storage & devices", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/settings/settings.go:572 'IsDefault bool `json:\"is_default,omitempty\"` // new apps use this by default' (line moved from 453); the default is honoured only in templates/deploy.html:614 '{{else if and .IsDefault (not .NotAllowed)}}selected{{end}}'; no internal/stacks caller of GetDefaultStoragePath/IsDefault.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/settings/settings.go"], "change": "Reword the field comment: the deploy FORM pre-selects this path (templates/deploy.html:614); the deploy API applies no default when HDD_PATH is omitted. (The alternative — a server-side default — is a behaviour change and not small.)", "test": "Comment-only; go vet ./internal/settings. Optionally a template render test asserting the default path is 'selected'.", "minutes": 15}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-371", "sev": "P4", "category": "Monitoring & notifications", "group": "NOT-WORTH-IT", "evidence": "Controller notifier events (internal/notify/notifier.go pushEventMsg list) include db_dump_completed, crossdrive_completed (:933) but no off-site success event; hub has no offsite *_completed type except offbox_abandon_completed.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The weekly off-site backup sends no 'done' event, while the two local tiers do. Failures and an 8-day staleness deadline are already alarmed.", "cost": "A new event type in the controller plus the hub allowlist and severity mapping (two repos), or a written decision that silence-on-success is intended.", "if_never": "The operator sees off-site success only through its absence of alarms and the backup card, as today.", "pick": "close-as-accepted with one line in 07-backup-architecture saying success is silent by design because failure and staleness are alarmed"}, "minutes_spent": 6}
|
||||
{"id": "R-372", "sev": "P4", "category": "Hub & operator", "group": "NOT-WORTH-IT", "evidence": "No hub or controller surface for 'never produced a copy' (grep 'never produced|NeverProduced' finds only an unrelated comment at felhom-controller/controller/internal/backup/backup.go:724). The underlying F-6E-1 was judged demo-data churn and the controller warns loudly.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "An optional idea from July: show 'this second-drive copy was never made because its source is missing' separately from 'last copy failed' in the operator screen.", "cost": "A design call on the operator view plus a new state carried controller->hub->UI.", "if_never": "The operator keeps seeing the existing loud warning; no data risk.", "pick": "close-as-accepted"}, "minutes_spent": 5}
|
||||
{"id": "R-373", "sev": "P4", "category": "Box system & updates", "group": "FIXED-BY-LATER-WORK", "evidence": "Premise (20G/50G two-volume mismatch, 'nothing sets SysDataGrowGB') was retired by agent v0.120.0 one-data-volume work, commit cd6e267 'v0.120.0 — one data volume (R-165...)'. felhom-agent/internal/reconcile/bringup.go:191 '// SysDataGrowGB is a COMPATIBILITY INPUT since agent v0.120.0 (R-165). There is no longer a second' and :437 'growGB := spec.DataVolGrowGB + spec.SysDataGrowGB'; installer passes it (felhom-host-install.sh:3140 '-sysdata-grow \"$SYSDATA_GROW\"') and records the sizing at :2041-2058. Note: cd6e267 predates the row's filing date; the row quoted an older audit.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-374", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "felhom.eu/documentation/audits/CAMPAIGN-12-class-sweep-2026-08-08.md:121-123 still says 'three of the 19 were called borderline and left unfiled' without naming them; the doc's only commit is b7fb2117 and no evidence file in audits/ lists the 19.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A July audit says three borderline cases were left unfiled but never named them.", "cost": "Re-running the whole C1 refusal sweep to re-find 19 cases and guess which three were meant — hours, and the guess cannot be checked.", "if_never": "Nobody can re-judge those three; the refusal class has had later sweeps and gates.", "pick": "close-as-accepted, with one line in the audit saying the three are not recoverable"}, "minutes_spent": 5}
|
||||
{"id": "R-375", "sev": "P4", "category": "Backup & restore", "group": "UNCHECKED", "evidence": "Requires a read-only check on ep0 (token's datastore audit permission); not verifiable from source and ssh is out of scope for this checker.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-376", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "Legend present ('not yet classified') in 00..06 and 10 of documentation/architecture/, but MISSING in the newer 08-alarm-ladder.md, 09-update-architecture.md and 11-os-updates.md (grep -ci 'not yet classified' = 0 each); 09 has zero [DESIGN]/[FACT] marks.", "dup_of": null, "unique_facts": "Three architecture docs added after the row (08, 09, 11) lack the legend.", "small_fix": {"repo": "felhom.eu", "files": ["documentation/architecture/08-alarm-ladder.md", "documentation/architecture/09-update-architecture.md", "documentation/architecture/11-os-updates.md"], "change": "Carry the same marker legend paragraph into the three documents written after the 2026-08-22 pass; then close the row, since 'mark as sessions touch them' is a standing practice already in the template, not a defect.", "test": "grep -ci 'not yet classified' returns >=1 in every numbered architecture doc; python3 scripts/repo_gates.py --fast.", "minutes": 15}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-377", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/CONTEXT.md:1537 '## Standing rulings' runs to EOF: 189,685 bytes, 0 '###' sub-headings, 153 bullets, 39 distinct S- ids; each ruling starts as a bold paragraph e.g. '**S-39 — \"WE DO NOT KNOW\" IS NEVER DRAWN AS \"FINE\"...'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["CONTEXT.md"], "change": "Turn each ruling's opening bold line '**S-NN — TITLE (date ...).**' into a '### S-NN — TITLE (date)' heading, changing no other byte; no compression, no reordering.", "test": "Scripted transform; assert 39 '### S-' headings and that git diff --word-diff touches only those opening lines (body bytes identical); repo_gates --fast.", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-390", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "Commit 2344589a ('... runbook pveam note'); felhom.eu/documentation/runbooks/RUNBOOK-manual-build.md:154 '2. Run **`pveam update` first** — the `virgin` snapshot's template INDEX is stale too, and a stale index fails as a bogus'.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-391", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog-felhom.eu/scripts/catalog_gates.py has no SHARED_ / observations entry (grep 'SHARED_|observations' returns nothing); app-catalog-felhom.eu/CLAUDE.md has no 'observation' mention.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["CLAUDE.md"], "change": "Take the row's second option: state in CLAUDE.md that the catalog REPORT.md carries no observations section by convention (findings go straight to the register), so gate 11 is not needed here. The runner refactor (first option) is the bigger alternative.", "test": "python3 scripts/catalog_gates.py --fast (instructions gate checks CLAUDE.md length) and python3 ../felhom.eu/scripts/check_skills.py unaffected.", "minutes": 15}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-392", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "ls felhom.eu/documentation/architecture shows no agent-tooling/workflow document (00-11 are all product; plus _design-review, _hub-review, _recovery-inventory). Writing a new architecture document is more than an hour and needs the operator's view of the split.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-393", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "No decision-log skill exists (felhom.eu/skills has 9 skills, none mention 'decision log'). Since then .claude/rules/unprompted-work.md §2 requires every unattended decision be recorded in CONTEXT.md + the owning architecture doc and put FIRST in the morning note (§4).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A proposed skill plus helper script to log every decision an unattended run makes.", "cost": "A storage convention, a helper, a skill and rotation rules — a small project with its own acceptance tests.", "if_never": "Unattended decisions are still recorded under the unprompted-work rule (CONTEXT.md + morning note); only minor in-run choices go unlogged.", "pick": "close-as-accepted (superseded in practice by unprompted-work.md §2/§4)"}, "minutes_spent": 4}
|
||||
{"id": "R-394", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "wc -l felhom.eu/skills/felhom-build-deploy/SKILL.md = 186 (was 179 at filing — grew); scripts/check_skills.py:45 GRANDFATHERED still holds the exemption. Trim needs a session that can verify the build/deploy commands it keeps.", "dup_of": null, "unique_facts": "The skill has grown from 179 to 186 lines since filing.", "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-402", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No hub Go/template reads last_integrity_ok/_depth (grep in hub/internal returns nothing); still allowlisted at felhom.eu/scripts/wire_contract_gate.py:163 and :220. Needs the operator's decision on what the screen says.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-416", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "Partly covered: scripts/register_shape_gate.py:120 'RULE 3 — duplicate: {rid} already has a row at line ...' (added in 462ab4a5, R-627) now refuses a duplicate id WITHIN OPEN-ITEMS.md only (REG path at :84 = OPEN-ITEMS.md). CLOSED-ITEMS.md has no within-file duplicate rule; closed_register_gate.py:53 still lists '4. A duplicate id WITHIN one register escapes.'", "dup_of": null, "unique_facts": "OPEN-ITEMS half is already fixed by register_shape_gate RULE 3 (462ab4a5); only CLOSED-ITEMS remains.", "small_fix": {"repo": "felhom.eu", "files": ["scripts/register_shape_gate.py or scripts/closed_register_gate.py", "scripts/test_gate_decoys.py"], "change": "Apply the existing RULE 3 duplicate check to CLOSED-ITEMS.md too (suffixed ids like R-88a/R-88b stay distinct), and update the closed_register_gate.py:53 hole list to point at it.", "test": "Planted-duplicate decoy in a temp CLOSED file must convict; suffixed-id control must pass; run against the live CLOSED-ITEMS.md first to see it is clean.", "minutes": 35}, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-418", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "It drifted AGAIN: felhom.eu/scripts/repo_gates.py docstring lists 1-14 (+9b) = 15 gates while GATES has 17 — 'script-tests' and 'decoy-coverage' are registered but not listed (python import: len(GATES)=17). No test compares the two.", "dup_of": null, "unique_facts": "Live drift right now: script-tests and decoy-coverage missing from the docstring.", "small_fix": {"repo": "felhom.eu", "files": ["scripts/repo_gates.py", "scripts/test_repo_gates.py"], "change": "Add the two missing gates to the docstring list and a test that parses the docstring's gate labels and asserts they equal [g[0] for g in GATES].", "test": "python3 -m unittest scripts/test_repo_gates.py; red-proof: the test fails on today's docstring before the two lines are added.", "minutes": 30}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-420", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "felhom.eu/scripts/repo_gates.py:106 tuple is (label, path, args, fast, exemptible) — still no 'blocking' field, as the row says; no gate there needs one.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The felhom.eu gate runner cannot mark a gate as advisory-only; the controller runner can.", "cost": "Add a field to the runner when a need appears.", "if_never": "Nothing today; no advisory gate is wanted in that repo.", "pick": "close-as-accepted (add it with the first gate that needs it)"}, "minutes_spent": 3}
|
||||
{"id": "R-421", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Deliberate class row ('stays open as the place the next instance is recorded'); its open instances R-422..R-426 are still open in this batch. Not a fixable item by itself.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-422", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/scripts/reuse_refs_check.py:41 PATH_RE still ends '\\.(?:go|py|html|css|yml|yaml|sh)\\b' — no .md. Widening it needs a false-positive walk across all four repos' REUSE.md/CLAUDE.md citations (the row says that pass is the work).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-423", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/site_gates.py:22-27 hardcoded PAGES list; website/ today holds exactly those 9 files, so nothing is missed today, but a new page is not scanned.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/site_gates.py", "scripts/test_gate_decoys.py", "scripts/decoy_coverage_gate.py"], "change": "Discover website/**/*.html and FAIL on any page not in PAGES (or glob and keep PAGES only as exceptions); flip the decoy test to expect conviction and drop the 'site' EXEMPT entry.", "test": "Decoy: a temp website/decoy-page.html must now convict; repo_gates --fast green on the real tree.", "minutes": 45}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-424", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "felhom.eu/scripts/one_register_gate.py:23 '...defect written under `idea` looks exactly like a proposal to this gate.' — hole declared in the gate itself.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The roadmap gate cannot tell a real defect filed as an 'idea' from a genuine idea.", "cost": "No mechanical fix exists; telling a defect from a proposal is human judgement.", "if_never": "A mis-filed defect could hide on the roadmap until a person reads it.", "pick": "close-as-accepted (hole stays declared in the gate's docstring)"}, "minutes_spent": 3}
|
||||
{"id": "R-425", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/scripts/offbox_rename_gate.py:16-20 FILES = backups.html, offbox_handlers.go, offbox.go only; templates/backups_remote.html exists and is not scanned, and customer copy now lives in internal/i18n/locales/hu.json (20 'NAS' hits) which is not scanned either.", "dup_of": null, "unique_facts": "Since localisation, the customer copy the gate guards moved to locales/hu.json, which FILES does not include — the gate may now be largely hollow, not only narrow.", "small_fix": {"repo": "felhom-controller", "files": ["controller/scripts/offbox_rename_gate.py", "controller/scripts tests (decoy)"], "change": "Scan by pattern (templates/backups*.html, *offbox*.go) plus internal/i18n/locales/hu.json, or assert FILES against a discovered set so an unclassified file fails; drop its decoy-coverage exemption.", "test": "Decoy: 'NAS-mentés' in a temp backups_offbox_extra.html and in a hu.json value must convict; real tree passes (check the 20 existing NAS strings are allowed forms).", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-426", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/scripts/decoy_coverage_gate.py EXEMPT now has 19 entries (loaded via python): hub-copy is gone, felhom-agent 'release-complete' is new; group (d) gates (hostinstall, wire-contract, due-checks, published, image-resolvable, volume-persistence) all still exempt.", "dup_of": null, "unique_facts": "Count is 19 now, not 20: hub-copy left the list; felhom-agent release-complete joined it.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-427", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "Commit 71b8c8c6 (Backlog triage Part B: '... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS'); felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word (CLOSED, SHIPPED,'. Of the 12 named rows, R-385/387/341/378/405/88a/88b/123 are now only in CLOSED-ITEMS.md; R-190 and R-352 remain open (partly-closed, as the row predicted).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-437", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu 71b8c8c6 'Backlog triage Part B: 125 finished rows + 20 id-less rows moved to CLOSED-ITEMS ... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS (decoys, seen red) ... register 444 -> 325'. felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word'. Gate run today: 'closed-register gate OK — no open work filed as closed, no id in both registers.' (456 closed / 336 open, 0 convicted).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-445", "sev": "P4", "category": "Hub & operator", "group": "NOT-WORTH-IT", "evidence": "Still true in mechanism: hub/internal/web/apps.go:102-106 computes SuggestedLimit from fleet P95 with no deployment-count/live check. But hub/internal/web/apps.go:67 'since := parsePeriod(period, 7*24*time.Hour)' — the detail page defaults to a 7-day window, so a throwaway's samples leave the default view within a week on their own; the 2026-09-01 sample is long outside it.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The hub's per-app memory suggestion can be built from samples of an app that no longer runs anywhere (e.g. a 15-minute test install).", "cost": "An operator ruling plus a hub change (age-out or exclude zero-deployment apps) and a test, ~1-2 h.", "if_never": "An operator glancing at the app page within 7 days of a throwaway install could see a suggestion based on it; after 7 days the default view drops it. Nothing customer-facing.", "pick": "close-as-accepted"}, "minutes_spent": 5}
|
||||
{"id": "R-451", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/internal/report/types.go ContainerDetailReport still carries only Name/State/CPUPercent/MemoryMB (no image field). Ruled (09 §3 decision 18), build deferred until fleet grows; needs controller payload + hub denormalisation + fleet page across two repos.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-454", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "The five files named are now clean (gofmt -l controller/internal/web/ prints nothing), but `gofmt -l controller` in felhom-controller lists 12 OTHER files today: cmd/controller/main.go, internal/agentapi/diskverdict.go, internal/api/update_reason_test.go, internal/appbackup/namespace_root_test.go, internal/appbackup/r381_undo_naming_test.go, internal/backup/r669_applied_meta_test.go, internal/family/family.go, internal/infra/infra.go, internal/notify/r636_oom_storm_test.go, internal/quiesce/tiers_test.go, internal/stacks/delete.go, internal/stacks/life_records.go. `grep -rn gofmt --include=*.py` in felhom-controller: no hit — controller/scripts/controller_gates.py still has no formatting gate. The row's 'the count can only grow' is proven.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/scripts/controller_gates.py", "the 12 files listed in evidence"], "change": "Add a gofmt -l gate (fails on any listed file, INCONCLUSIVE if gofmt missing) to controller_gates.py, see it red on today's tree, then one gofmt -w formatting commit for the 12 files.", "test": "Run controller_gates.py before (red, lists 12) and after (green); go build ./... and go vet unchanged.", "minutes": 30}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-457", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "No later commit references R-457 beyond the filing release (felhom-controller 38d28b5 v0.234.0 / 998aa31 REPORT). No faked-clock CI run exists (grep for faketime/FAKE_NOW in felhom-controller: no hit). The six candidate files are named but unread.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["internal/backup/offbox_test.go", "internal/web/handler_export_upload_test.go", "internal/web/r103_tier2_action_test.go", "internal/web/dashboard_backup_card_test.go", "internal/web/async_restore_test.go", "internal/stacks/installed_test.go"], "change": "Read the six candidates; for each date literal that feeds an assertion evaluated against time.Now(), derive it from now (as the R-457 fix did). The faked-future-date CI instrument is a separate, larger idea and should be split out or dropped.", "test": "Run the touched packages' unit tests that do not reach Docker (or rely on CI); record per file 'literal feeds real-clock assertion: yes/no'.", "minutes": 60}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-460", "sev": "P4", "category": "App updates", "group": "NOT-WORTH-IT", "evidence": "Fact about the BookStack app (API token mintable only via web UI; secure cookies over plain http give 419), not a code defect; re-measured 2026-09-21 per row. DB half proven by app-catalog-felhom.eu/scripts/upgrade-test.py.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "BookStack's uploaded files cannot be checked automatically after an upgrade; only its database can.", "cost": "An upstream headless token route, or a browser-driven step DooPlex cannot run.", "if_never": "BookStack upgrades stay half-proven automatically; file survival rests on volume persistence and manual checks.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-464", "sev": "P4", "category": "App updates", "group": "FIXED-BY-LATER-WORK", "evidence": "Lesson homed and harness uses the correct probe. app-catalog-felhom.eu b7ef0c4 'upgrade-test.py: record the engine's own view of its datadir'; app-catalog-felhom.eu/scripts/upgrade-test.py:278 '\"mariadb-upgrade --check-if-upgrade-is-needed --user=root \"'. felhom.eu d6837d98 (SPIKE R-459); felhom.eu/documentation/architecture/09-update-architecture.md:1647 '1. **Ask the engine, not the log.** MariaDB's entrypoint prints `MariaDB upgrade not required` on an' (cites R-464).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-488", "sev": "P4", "category": "Process & tooling", "group": "UNCHECKED", "evidence": "The claim is a measured suite runtime (5.5 min); confirming it needs running go test ./internal/backup, which I did not run (read-only; backup tests may reach real docker on DooPlex). No commit after filing (felhom-controller 24d7c54) mentions R-488 or a test-speed change in internal/backup (git log --grep on internal/backup since 2026-09-13: empty), so it is likely still true.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-492", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "cfg.Paths.HDDPath still defined and read: controller/internal/config/config.go:167 'HDDPath string `yaml:\"hdd_path\"`', :453 envStr(\"FELHOM_PATHS_HDD_PATH\", &cfg.Paths.HDDPath); readers internal/report/builder.go:69, internal/monitor/healthcheck.go:79, internal/api/router.go:1135, internal/web/server.go:988, cmd/controller/main.go:358 and :526. Only 2 test references (1 file).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/config/config.go", "controller/internal/report/builder.go", "controller/internal/monitor/healthcheck.go", "controller/internal/api/router.go", "controller/internal/web/server.go", "controller/cmd/controller/main.go (gitignored dir — use git add -f)", "settings.AutoDiscoverStoragePaths signature"], "change": "Remove Paths.HDDPath, its env binding and each reader's dead global branch, keeping the per-app/discovered fallbacks each reader already uses.", "test": "go build ./... proves no reader remains; existing tests of the six readers stay green; a config test that FELHOM_PATHS_HDD_PATH is no longer read.", "minutes": 60}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-494", "sev": "P4", "category": "Install & onboarding", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub/internal/cloudflare/ holds only unblock.go; no tunnel/DNS creation code in hub (grep cfd_tunnel: none). Building it is a new Cloudflare-API mechanism on the hub; operator ruled it non-blocking.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-501", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu a4993272 'CLAUDE.md: the CI-check recipe was wrong in two ways, both measured today'. felhom.eu/CLAUDE.md:176 'rows — a run can sit several pages earlier. **Scan every page** and match on `head_sha`; with a'; recipe at CLAUDE.md:166-168 loops every page.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-502", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/iso/test/bootstrap-modes.sh exists; `grep -rn 'bootstrap-modes|bootstrap_modes' scripts/*.py .gitea/workflows` in felhom.eu returns nothing — no gate or CI runs it.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/repo_gates.py", "scripts/iso/test/bootstrap-modes.sh"], "change": "Register bootstrap-modes.sh in repo_gates.py behind a docker-available check that reports INCONCLUSIVE (never pass) when docker/the felhom-iso-assistant image is absent, plus a decoy (a broken banner must turn it red).", "test": "Run repo_gates.py with docker present (green), with the harness sabotaged (red), and with docker hidden from PATH (INCONCLUSIVE).", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-503", "sev": "P4", "category": "Install & onboarding", "group": "NOT-WORTH-IT", "evidence": "Not built (by design); the 2026-07-31 ruling 'a person chooses the disk' stands per the row; no later commit references R-503 other than its filing context.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "An idea, offered and not chosen: the installer would pick the disk itself when there is exactly one.", "cost": "A spike with three measurements and then reversing two operator rulings.", "if_never": "Nothing changes: a person keeps choosing the disk, as ruled.", "pick": "close-as-accepted (as DECLINED by ruling; the three measurements stay in the closed row for any future reversal)"}, "minutes_spent": 2}
|
||||
{"id": "R-504", "sev": "P4", "category": "Install & onboarding", "group": "UNCHECKED", "evidence": "The claim (iso.felhom.eu/ returns 404) is live-only; I may not curl hosts. felhom.eu/documentation/runbooks/VOLUNTEER-first-hour.md:14 still says '`iso.felhom.eu/` itself still has no index — R-504'. Fix needs a Cloudflare rule the operator owns. Cosmetic; households use felhom.eu/letoltes (website/letoltes.html exists).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-507", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Needs measuring QEMU input-send-event or a VNC client against a live VM on felhom-pve — a live-machine spike, not a source change. No later commit references R-507.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-525", "sev": "P4", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Row itself states it is a new unmeasured mechanism (forwardAuth / Quantum proxy auth) needing a scratch-guest spike.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-526", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub/internal/tenantsync/client.go:110 '// Deprovision DESTROYS the customer's PBS namespace, all its backup groups, and its token — the'; no token-only op exists. Needs a new op on protected ep0 and an operator yes/no.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-527", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "Row's wording is partly WRONG: the flag IS read — controller/internal/stacks/deploy.go:401 'if field.LockedAfterDeploy {', :790-791 and :1350-1374 copy it into appCfg.LockedFields, and deploy.go:710 UpdateStackConfig refuses a locked key. But UpdateStackConfig (deploy.go:681) has NO non-test caller (grep: only its own definition/logs), and deploy.html:508-518 renders every auto field `readonly`, so the effect the row describes (every field read-only after install regardless of the flag) is true.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A catalog flag that marks some settings 'locked after install' changes nothing visible: the page makes every setting read-only anyway, and the edit path that would honour the flag is never called.", "cost": "Either delete the flag from the catalog and controller (catalog-wide edit + parity fixtures), or build an edit-after-install feature (a product change).", "if_never": "Nothing a household meets: all settings stay read-only, which is the safe direction.", "pick": "close-as-accepted (with the corrected facts: flag read into LockedFields, enforced only by uncalled UpdateStackConfig)"}, "minutes_spent": 8}
|
||||
{"id": "R-532", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "app-catalog-felhom.eu/templates/vaultwarden/docker-compose.yml:30 '- SIGNUPS_ALLOWED=${SIGNUPS_ALLOWED:-false}' (image vaultwarden/server:1.36.0-alpine, :22). The /api/config field is upstream Vaultwarden behaviour; the server refusal (400) is live-only and not re-measured here.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Vaultwarden's web page shows a sign-up form even though sign-ups are off; the server then refuses it.", "cost": "An upstream fix, or a catalog note; nothing Felhom can change in the server's answer.", "if_never": "A stranger who finds the page sees a form that fails; an invited household is not affected.", "pick": "close-as-accepted"}, "minutes_spent": 3}
|
||||
{"id": "R-541", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Row: needs a new copy/re-key/release mechanism and a design; no later commit references R-541. Far off per re-rank (0.3% full, one pool box).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-544", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/hub/internal/web/hosts.go:917 's.logger.Printf(\"[INFO] host deleted: %s (escrow deleted: %v)\", hostID, deleteEscrow)'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/web/hosts.go", "hub/internal/web/hosts_delete_test.go"], "change": "Change the log line to state the effect, e.g. 'host deleted: %s (escrow custody demoted to retained)' when escrow existed, and drop the boolean name from the text.", "test": "A test in hosts_delete_test.go capturing the logger output on an acknowledged delete asserts 'demoted to retained' and the absence of 'escrow deleted'; red-proof against today's line.", "minutes": 30}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-551", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "Row: behaviour proven by five red-proofed ServeHTTP tests; live proof needs a box in the 'paused AND agent-connected' state, which only a fresh install or a local-API token on scratch guest 9202 (a live-machine change) gives. No later commit references R-551.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The escrow 'waiting for the agent' screens are proven by tests but never seen on a real box in that state.", "cost": "Provisioning a local-API token on the scratch guest, or walking it during the next fresh install.", "if_never": "Small risk the live page differs from the tested page; the state lasts ~17 minutes after a bind.", "pick": "close-as-accepted (add 'walk R-546 readiness branches' to the next fresh-install checklist instead)"}, "minutes_spent": 2}
|
||||
{"id": "R-555", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/wire_contract_gate.py:451 'def receiver_tokens(repo_root):' tokenises whole files: line 482 'toks.update(TOKEN_RE.findall(fh.read()))' — no comment stripping.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/wire_contract_gate.py"], "change": "Strip Go // and /* */ comments and template {{/* */}} before TOKEN_RE in receiver_tokens; add a decoy 'tag named only in a receiver comment must convict'. Newly surfacing tags each become a finding (allowlist with reason or a row).", "test": "Run the gate: decoy red, real tree result reviewed; the 'language' allowlist entry still justified.", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-564", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/scripts/retrieval_promise_gate.py:54 'STEMS = [\"visszaállíthat\", \"visszaszerezhet\", \"visszahozhat\", \"visszanyit\"]' — joined forms only, no split-verb pattern.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/scripts/retrieval_promise_gate.py"], "change": "Add split-form Hungarian patterns (állíthatók? vissza, (hoz|szerez|nyit)\\w* vissza), register the Hungarian occurrences found (the seven already reviewed in English), and add a planted split-verb decoy.", "test": "Gate red on the decoy, green on the tree after registration; search with ASCII fragments plus positive/negative controls per the Hungarian-search rule.", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-567", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "controller/internal/web/templates/layout.html:83 '{{$storageOpen := or (eq .Page \"storage\") (eq .Page \"storage-network\")}}' — storage_init/storage_attach not included; storage_handlers.go:351 'data := s.baseData(tmpl, title)' passes the template name as Page.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/templates/layout.html", "controller/internal/web/testdata/i18n_parity (re-capture storage_init/attach fixtures)"], "change": "Add (eq .Page \"storage_init\") (eq .Page \"storage_attach\") to $storageOpen and mark the Meghajtók link active for them.", "test": "Render test: GET /storage/init and /storage/attach carry 'nav-group is-open' and the active storage link; red before, green after; re-capture parity fixtures.", "minutes": 40}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-568", "sev": "P4", "category": "Storage & devices", "group": "STILL-TRUE-SMALL", "evidence": "controller/internal/web/disk_health.go:124-152 diskHealthRows appends rows in resp.Disks order; no sort in the file (grep 'sort.' in disk_health.go: none).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/disk_health.go", "controller/internal/web/disk_health_test.go"], "change": "Sort rows by diskKey(d) (durable id, falling back to name) before returning.", "test": "Unit test feeding the fake agent two disks in both orders and asserting one identical row order; red-proof by removing the sort.", "minutes": 30}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-569", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "controller/internal/api/router.go:742 'if strings.Contains(err.Error(), \"protected\") {', also :745, :1018, :1021, :1024 ('not deployed'/'still running'), :1102, :1105, :1108 ('not orphaned').", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/api/router.go", "controller/internal/stacks (sentinel errors)", "router tests"], "change": "Add KindErrorf-style sentinels in internal/stacks (protected, not found, not deployed/still running, not orphaned), a statusFor helper per handler family replacing the three Contains blocks.", "test": "One table test per family passing a REWORDED error message and asserting the same status code; red against today's string matching.", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-570", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Fallback still present: controller/internal/web/handlers.go:1050 'offboxStaleWarningMarker = \"nincs mentésre jelölt alkalmazás\"'; producer internal/backup/offbox.go:1168. Closing depends on a fleet condition (every box one off-site run on >=0.251.0) — a watch, not a fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-571", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "grep ClassifyOffsiteFailure|PageOnly|Inline in felhom.eu/documentation/architecture/07-backup-architecture.md and 02-controller-module-map.md: no hit. Classifier lives at felhom-controller/controller/internal/backup/offbox.go:193 'func ClassifyOffsiteFailure(err error) OffsiteFailureClass {'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["documentation/architecture/07-backup-architecture.md", "documentation/architecture/02-controller-module-map.md"], "change": "Add a short section to 07 listing the six failure classes, what each means for the customer, and that restic/ssh signatures are external; add an alert-placement paragraph (inline under storage bars vs top banner) to 02.", "test": "Doc-only: grep the two docs for ClassifyOffsiteFailure/PageOnly afterwards; the repo's doc gates stay green.", "minutes": 40}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-574", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "controller/internal/web/handler_debug.go still carries 40 lines with accented Hungarian string literals (grep -cP count); last touched by 0c702f8 v0.279.0, not converted. Labelling ~40 literals page-copy vs payload, adding en/hu keys and parity fixtures exceeds an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-576", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/scripts/i18n_go_parity.py has no call-site argument-count or '+'-adjacency check (grep verb/argument: only VERB_RE/strip_verbs for text equality, lines 76-78, 196-199). Parsing multi-line Go call arguments reliably from Python with decoys is likely >1 h.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-577", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Waiting on an operator decision (what the share feature promises a stranger); felhom.eu/documentation/architecture/10-localisation.md table row 'the two guest share pages, the catch-all | a stranger / nobody | **no globe** | — (R-577, the operator's)'.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-579", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Gate deliberately deferred until R-554 deletes the first-boot wizard; R-554 is still OPEN (OPEN-ITEMS.md:131). Versionless links remain in controller/internal/setup/templates/setup_*.html:8 '<link rel=\"stylesheet\" href=\"/static/style.css\">' (8 files).", "dup_of": null, "unique_facts": "NEW: controller/internal/web/templates/monitoring.html:255 '<script src=\"/static/chart.min.js\"></script>' also has no ?v= — outside the wizard, so the future gate must cover or allowlist it. Could be folded into R-554's closing work.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-588", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/documentation/runbooks/iso-release-gate.md:319-322 'Result recording' says only 'in the release report' — names no home. documentation/tests/ holds iso-release-1.27.0, 1.27.1, 1.29.0 only; no record dir for 1.28.0 or later ISOs.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["documentation/runbooks/iso-release-gate.md"], "change": "Name the single home documentation/tests/iso-release-<ver>-<date>/ in the Result-recording section, and add a pointer dir/README for 1.28.0 to its audit record (evidence-backup-promise-2026-09-16/phaseD-iso-gate.txt). Optional existence check left out.", "test": "Doc-only; the repo doc gates stay green.", "minutes": 20}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-591", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "The copy is deepCopyStack in controller/internal/stacks/manager.go (row says Copy()); it deep-copies AppConfig (:1137-1148), DeployFields (:1162), OptionalConfig (:1174), Integrations (:1186) and has no I18n line (grep I18n in manager.go: none). Meta.I18n defined at internal/stacks/metadata.go:94.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/stacks/manager.go", "a stacks unit test"], "change": "Deep-copy Meta.I18n in deepCopyStack (map plus nested values).", "test": "Test mutates the copy's I18n overlay and asserts the original is unchanged; red without the copy.", "minutes": 30}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-594", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog-felhom.eu/scripts/check-copy-i18n.py: grep ALLOWLIST/allowlist — no hit; no way to register a true occurrence.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["scripts/check-copy-i18n.py"], "change": "Add ALLOWLIST_EN of (app, path, reason); registered occurrences pass, unregistered convict, and any entry matching nothing is itself a failure.", "test": "Decoys: a registered occurrence passes, an unregistered one convicts, a stale entry convicts. Mind catalog CI has no PyYAML (degraded mode must still run).", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-599", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/hub/internal/web/hosts.go:898 'http.Error(w, \"Host is ONLINE — deletion is refused (a live agent would receive 401s permanently).\", http.StatusConflict)' — no last-report age or opening time. target-selection.md: no mention of the stale-threshold wait (grep 45 min|stale_threshold: none).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/web/hosts.go", "hub/internal/web/configs.go (config delete 409)", "documentation/runbooks/target-selection.md"], "change": "Make both 409 bodies say when the last report arrived and when deletion opens (last report + configured stale threshold), and name the wait in target-selection.md's drill section.", "test": "Handler test with a host last reported N minutes ago asserts the 409 body carries the minutes and the opening time computed from the configured threshold (not the 30m literal).", "minutes": 60}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-602", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu e02bc038 'hub v0.119.0 — ... R-596/R-598 closed' added the finding; felhom.eu/documentation/architecture/10-localisation.md:809-812 'the `felhom_lang` cookie and got the **Hungarian** page for `en`. ... The cookie is the right instrument for the anonymous claim page and the **wrong**' and :509 '`langFor`'s order is fixed: `?lang=` → **the household's setting when a session exists**'. Only the optional pointer from the workspace live-validation rules is absent (grep ?lang=/felhom_lang in CLAUDE.md files and .claude/rules: none) — a 5-minute add if wanted.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-603", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "No gate or helper: grep for html.EscapeString(want|'|R-603 in felhom-controller/controller scripts+internal: none. controller/internal/i18n/locales/en.json has 27 lines containing an apostrophe today, so the trap is live for any Contains assertion on those.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["a test helper in internal/web (e.g. assertPageContains)", "optionally controller/scripts gate"], "change": "Add a test helper that compares against html.EscapeString(want) and use it in the render tests that assert English copy; the bundle gate with a 27-entry allowlist is the larger alternative.", "test": "A test asserting an apostrophe-bearing en.json value through the helper passes; the same via raw strings.Contains fails (red-proof).", "minutes": 45}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-605", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog-felhom.eu/scripts/catalog_gates.py:122 'VERDICT = {0: \"OK\", 1: \"FAILED\", 2: \"INCONCLUSIVE\"}' — harness refusal and per-app undetermined both exit 2 and print the same word.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["scripts/catalog_gates.py", "scripts/check-volume-persistence.py", "scripts/check-image-resolvable.py"], "change": "Give harness-level refusal its own exit code (e.g. 3 = REFUSED) in both scripts and map it to a distinct label in catalog_gates.py.", "test": "Decoys each way: a forced canary failure reads REFUSED, a per-app undetermined reads INCONCLUSIVE.", "minutes": 60}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-610", "sev": "P4", "category": "App updates", "group": "NOT-WORTH-IT", "evidence": "felhom-controller@7690c27 controller/internal/stacks/update.go:1453 `case UpdatePhaseStarting, UpdatePhaseVerifying:` - starting and verifying share ONE recovery arm, the one measured three times. No fault injector exists (grep for faultinject/failpoint in controller/ returns nothing).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A power cut landing inside the sub-second `starting` phase has never been measured; all three live cuts landed in `verifying`, which runs the same recovery code.", "cost": "An in-process fault injector in the controller (new mechanism) plus a live drill on a scratch guest.", "if_never": "Nothing new is learned: the code path is the same branch already proven three times; the residual risk is a difference between the two phases that the source does not show.", "pick": "close-as-accepted"}, "minutes_spent": 5}
|
||||
{"id": "R-617", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu@462ab4a5 (2026-09-22) documentation/architecture/09-update-architecture.md:1969 \"`POST /api/v1/repos/migrate` is the route that works (the project's Gitea tokens carry\" - continues at :1970 \"`write:repository` but not `write:user`, so `POST /user/repos` answers 403\"; recipe at :1979. The one-line note the row asked for exists (in the architecture doc rather than operations/). The optional operator-scoped token is a separate wish, not the defect.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-618", "sev": "P4", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Remaining open work = the controller-side idea (let `verifying` accept docker's own `healthy`). grep for docker health status in felhom-controller@7690c27 controller/internal/stacks/update.go returns nothing; no R-618 reference in Go source. It is an undecided design question (operator), not a defect; the three probe fixes (app-catalog@793c4fb) and the gate scripts/check-probe-matches-compose.py are done.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-619", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/api/router.go:395 `meta, appCfg, err := r.stackMgr.GetDeployFields(name)` then :402 `\"metadata\": meta,` - metadata served verbatim, no derivation of Required for password; deploy.go:366 \"// We never silently auto-generate — the user needs to know their password.\" still refuses.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": "controller/internal/api/router.go (getDeployFields), new test in controller/internal/api/", "change": "In getDeployFields, copy meta.DeployFields and set Required=true for every field with Type==\"password\" before writing the response (copy, do not mutate the shared metadata; the web deploy page uses GetDeployFields separately and is untouched).", "test": "Go unit test: a stack whose .felhom.yml has a type: password field with required: false; GET /api/stacks/<n>/deploy-fields must answer required:true for it and leave a secret field's required as declared; red-proof by running it before the change.", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-621", "sev": "P4", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Capture is done: felhom-controller@7690c27 controller/internal/stacks/update.go:1054 `outDir := filepath.Join(dir, \"hold-logs\", ts)`. The open part (show it on the app page's hold panel / logs fallback) is not built: grep for hold-logs/holdLogs in controller templates and handlers returns only update.go:1050-1082 and undo.go:535,550 (writers). Surfacing needs a page change with HU/EN copy and a design for which hold's log to show.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-624", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Row's own latest update: remaining class is vaultwarden (closed sign-up by design) and code-server; the open decision is whether the harness may hold an app's admin secret (operator). Not verifiable further from source; needs a decision, not a fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-644", "sev": "P4", "category": "Apps & catalog", "group": "UNCHECKED", "evidence": "About the live state of gokapi on scratch guest 9202 (crash-loop, deployed:true). Only the box shows it; no ssh allowed.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-652", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog@917a779 templates/romm/.felhom.yml:249 still carries `\"memory_peak_pct\": 80.9` with `\"memory_tight\": true` and no `memory_basis: anon` (contrast paperless-ngx/.felhom.yml:249 `\"memory_basis\": \"anon\"`). Needs a live re-measure of romm plus an undecided cache-thrash rule.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-654", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "app-catalog@917a779 templates/opengist/docker-compose.yml:11 `image: ghcr.io/thomiceli/opengist:1.15`; grep for \"/-/\" in templates/opengist/.felhom.yml returns nothing - app_info does not mention the moved pages.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "opengist 1.15 moved its pages under /-/; an old /login bookmark answers 404. The front page redirects correctly.", "cost": "A copy decision by the operator plus a HU/EN app_info line.", "if_never": "A household with an old deep bookmark sees a 404 once and re-bookmarks from the front page.", "pick": "close-as-accepted"}, "minutes_spent": 3}
|
||||
{"id": "R-687", "sev": "P4", "category": "App updates", "group": "NOT-WORTH-IT", "evidence": "Gaps (1)-(3) are unit-tested only (as the row says); item (4) proven live 2026-09-30. Cosmetic text still as described: felhom-controller@7690c27 controller/internal/quiesce/quiesce.go:846 `return true, fmt.Sprintf(\"the automatic update leg is running (it starts no step after %s)\", backupwindow.FmtHHMM(mod1440(startMin+stopMin)))` - derived from the window, not a manual leg's own deadline.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Three live proofs a scratch box cannot give (a 3-hour leg, a failing off-site leg, a files_may_change step without a whole copy) plus one log text that names the window's deadline instead of a manually started leg's deadline.", "cost": "Each live gap needs a special venue (off-site target, long leg); the log text is ~30 min but only matters for the manual debug chain.", "if_never": "The unit tests remain the proof; an operator reading the manual-chain log sees a deadline 20 minutes off.", "pick": "close-as-accepted"}, "minutes_spent": 5}
|
||||
{"id": "R-688", "sev": "P4", "category": "Hub & operator", "group": "NOT-WORTH-IT", "evidence": "felhom.eu hub/internal/web/customer_delete.go:130 \"// R-688 (v0.125.0): NO leg of the cascade calls Cloudflare. The dialog used to promise\" and :133 `cfManual := cloudflareManualRemoval(cfg)` - the dialog lists the manual removal; no Cloudflare leg exists.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Deleting a customer does not remove their Cloudflare tunnel and DNS records; the dialog now says so and lists what to remove by hand.", "cost": "A new Cloudflare leg in the reset cascade (API calls, ids, failure handling, tests) - a new mechanism touching an external service.", "if_never": "The operator removes a tunnel and a few DNS records by hand at each customer delete, guided by the preview.", "pick": "close-as-accepted"}, "minutes_spent": 4}
|
||||
{"id": "R-691", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Open work = the Tier 2 (second-drive) path of Use/Load is not live-proven; needs a two-drive Tier-0 box (9202 has one drive). Live-only gap, not a source defect; not checkable from source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-693", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "grep for memory_scales_with_limit / scales_with_limit across app-catalog-felhom.eu, felhom-controller/controller and felhom.eu/scripts returns nothing - no basis that tells growth-to-fill from pressure exists. Needs a design (new harness signal).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-705", "sev": "P4", "category": "Process & tooling", "group": "FIXED-BY-LATER-WORK", "evidence": "The remaining half (manual whole-guest backup) EXISTS and predates the row: felhom-controller bbed5af (v0.47.0, 2026-06-12) 'backups page — whole-guest backup visibility + manual trigger'. Live source @7690c27: controller/internal/web/backup_handlers.go:324 `case r.URL.Path == \"/api/guest-backup/trigger\" && r.Method == http.MethodPost:`; :340 `if err := s.backupTrigger.TriggerNow(); err != nil {`; quiesce.go:427 \"manual backup requested — quiescing now\" (bypasses due-ness, all tiers); wired cmd/controller/main.go:2099 and the page button backups.html:229. Agent side: felhom-agent internal/localapi/server.go:514 `mux.HandleFunc(\"POST /backup\", ...)`. The controller half was built v0.279.0 (night-chain, handler_debug.go:79). The row's 'no manual trigger' claim was not true at writing; it is not chained into night-chain, which the row did not require.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-707", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Open work = live proof that the gate OPENS for seerr, outline and rallly (needs a media server / e-mail on a test box). Live-only proof gap; the gating itself is in source (catalog 6faf432 per row).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-718", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller@7690c27 controller/internal/web/templates/app_info.html:112 `<p>{{T \"app_info.close_signup_text\"}}</p>` - the close card has no restart line, while the window card has one only at :132-133 `{{- if .SignupNative}}` / `{{T \"app_info.signup_window_restart\"}}`. en.json:2505 close_signup_text says nothing about a restart.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": "controller/internal/web/templates/app_info.html, controller/internal/i18n/locales/hu.json + en.json, the app-info handler (if SignupNative is not set when CloseSignupOffered)", "change": "Add a key app_info.close_signup_restart (\"The app restarts once for this.\" / HU twin) and render it under the close card when the app has an after_setup env switch (the same SignupNative fact); add the same sentence to the gate-open confirmation where after_setup.env exists.", "test": "Render test per branch: app with after_setup.env shows the line on the close card, app without it does not (red first); run the design-v2 copy/i18n gates.", "minutes": 60}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-719", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Built: felhom.eu hub/internal/web/selfbind.go:160 \"// R-719 (v0.126.0): „Új linket kérek\" on an expired or used link.\" Open part is the operator's review of the changed shape and a live mint+send proof (unit-proven only) - an operator decision, not a CC fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-725", "sev": "P4", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu hub/internal/i18n/locales/hu.json:79 `\"bind.invalid.body\": \"A hivatkozás 7 napig érvényes. Ha lejárt, kérj újat az ügyfélszolgálattól, ...\"` still sits above hu.json:87 `\"bind.resend.button\": \"Új linket kérek\"`; en.json:79 'ask support for a new one'. The console 'V' is the ✔ glyph in scripts/iso/felhom-bootstrap.sh:99 `printf ' Felhom — a doboz össze van kötve. ✔\\n\\n'` (golden scripts/iso/test/golden/bound.hu.txt:2). The English JSON to phone apps is left on purpose.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": "hub/internal/i18n/locales/hu.json, hub/internal/i18n/locales/en.json (bind.invalid.body); optionally scripts/iso/felhom-bootstrap.sh:99 + scripts/iso/test/golden/bound.hu.txt", "change": "Reword bind.invalid.body to point at the button below (e.g. 'Ha lejárt, kérj újat lent.' / 'If it has expired, ask for a new one below.'), keeping the operator alternative; optionally drop the ✔ glyph the console font renders as 'V' and update the golden.", "test": "Hub i18n parity/copy tests and a render test of the expired bind page asserting the new sentence and absence of 'ügyfélszolgálattól'; ISO golden test if the glyph is changed. Ships with the next hub release.", "minutes": 30}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-731", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "The shape-switch control lives only in audit tools: felhom.eu/documentation/audits/catalog-currency-2026-09-30/00-currency.py, 04-analyse.py; no standing currency script in app-catalog-felhom.eu/scripts or felhom.eu/scripts (ls/grep for currency/shape returns only golden_currency_gate.py, which is unrelated). Making it standing means promoting a registry-reading tool with tests - more than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-734", "sev": "P4", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "grep for '.immich' / hash ignore list in app-catalog-felhom.eu/scripts/*.py and templates/immich/.felhom.yml returns nothing - no exclusion exists. The row says the rule change needs an operator word; calibre-web shows the mark is sometimes right, so the rule needs design.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-739", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog@917a779 templates/wanderer/docker-compose.yml:117 `image: getmeili/meilisearch:v1.36.0`; grep MEILI_UPGRADE_DB in the compose returns nothing. Remaining: the template switch, a fixture (PocketBase create refused) and a measured step on the bench - live work.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-759", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Five checklist rows of wger need live measurement on 9202 (2.5, 3.7, 6.3, 8.2, 9.1); not verifiable from source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-760", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog@917a779 templates/vikunja/docker-compose.yml: service vikunja (image: vikunja/vikunja:2.6.0, line 12) has no `healthcheck:` key and no comment explaining why (whole file read).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": "templates/vikunja/docker-compose.yml", "change": "Read the vikunja 2.6.0 image config for a HEALTHCHECK/shell; if none and the image has no shell, add a comment saying why there is no compose healthcheck (like adventurelog-frontend's R-655 comment); otherwise add a healthcheck of the family the image supports (REUSE.md §2). No image: line moves, so no catalog_since.", "test": "scripts/onboarding_gaps.py row 4.1 no longer lists vikunja (or lists it as explained); catalog_gates.py --fast green; if a healthcheck is added, a deploy on 9202 must read healthy.", "minutes": 45}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-761", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog@917a779 templates/paperless-ngx/.felhom.yml:22 `# Logo: {assets.base_url}/assets/{slug}-logo.webp` vs felhom-controller controller/internal/config/config.go:511 `return fmt.Sprintf(\"/static/assets/%s-logo.svg\", slug)` and :516 `-logo.png`.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": "templates/paperless-ngx/.felhom.yml (comment block lines 20-24)", "change": "Change the comment to name `{slug}-logo.svg` (preferred) and `{slug}-logo.png` (fallback), matching config.go AppLogoURL/AppLogoPNGURL; comment-only.", "test": "grep shows no '-logo.webp' in the template comment; catalog_gates.py --fast green (copy-freeze gates ignore comments).", "minutes": 10}, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-764", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "grep smtp/mail in app-catalog templates/wger/.felhom.yml and docker-compose.yml finds only first_steps text (.felhom.yml:74 'Add meg az email címedet a beállításokban'); no smtp_mapping. A mapping needs a live boot proof with mail off (REUSE.md §2) - more than an hour; wger is hidden.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-766", "sev": "P4", "category": "Apps & catalog", "group": "FIXED-BY-LATER-WORK", "evidence": "Hub releases after the assets push (felhom.eu 40f07429, 2026-10-01): hub v0.131.0 (2026-10-04) .. v0.136.0 (d4be9f6f, 2026-10-05). The build copies website assets: felhom.eu scripts/build-hub.sh:98 `cp \"${WEBSITE_ASSETS_DIR}\"/*-logo.svg \"${BUILD_DIR}/assets/\" 2>/dev/null || true`, and the hub build workspace /mnt/5_hdd/felhom.eu/build/felhom-hub/workspace/assets/ holds radicale-logo.svg + 3 screenshots (also karakeep, dawarich) dated Oct 5 14:28; hub/Dockerfile:27 `COPY assets/ /usr/share/felhom/assets-seed/`. Not checked: what a live box shows (no machine access).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-768", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "Fit-check decline recorded in felhom.eu/documentation/audits/new-apps-2026-10-01/FIT.md (per row); nothing in source to fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Grimoire is not built because upstream rules out public exposure and ships no image for v1.x; the row only watches for that to change.", "cost": "A periodic re-read of upstream at each catalog campaign.", "if_never": "Nothing; Karakeep covers bookmarks. The re-read can live in the next catalog campaign's checklist instead of an open row.", "pick": "close-as-accepted"}, "minutes_spent": 1}
|
||||
{"id": "R-769", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "New-app idea waiting on an operator decision (a fork as new upstream, after R-767). No source to check.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-770", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "New-app idea waiting on the operator's go/no-go (CC recommends not building). No source to check.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-771", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "New-app idea waiting on the operator's go/no-go. No source to check.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-779", "sev": "P4", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Live proof gap on the real Cloudflare tunnel needing the operator's phone off wifi; not checkable from source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-781", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog@917a779 scripts/test_gate_decoys.py:498 `sh([\"git\", \"clone\", \"-q\", \"file://\" + ROOT, cat], cwd=ROOT)` clones the REAL records while the stand-in sibling holds only :502-505 `proof.txt`; real records cite sibling evidence (grep -c 'felhom.eu/documentation': onboarding/karakeep.md 45, dawarich.md 43, radicale.md 40). Last change to the file (96829d0, 2026-10-02) added family-gate cases only. Test not run here (it can reach a registry).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": "scripts/test_gate_decoys.py (onboarding_cases)", "change": "After the clone, delete the real onboarding/<app>.md records (all but _TEMPLATE.md and the exempt wger.md the cases use) from the scratch clone, so genuine cases judge only the records they build; alternative: point the stand-in sibling at the real felhom.eu documentation tree read-only.", "test": "Run scripts/test_gate_decoys.py: the four genuine onboarding cases go from FAIL to ok, all decoys still refused; the real check-onboarding gate stays green.", "minutes": 30}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-786", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog@917a779 onboarding/sparkyfitness.md has 8 '| open' rows, incl. :18 0.5, :20 0.7, :28 1.6, :29 1.7, :58 5.4, :68 8.3 (plus 0.1 licence = R-784, 2.1 = R-807). Most need live measurement (runtime internet, phone sign-in, second memory watch).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-793", "sev": "P4", "category": "Business & legal", "group": "NOT-WORTH-IT", "evidence": "Watch item from felhom.eu/documentation/audits/licences-2026-10-02/TABLE.md; nothing wrong as the catalog runs them (row's own reading).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Four apps ship enterprise/BUSL code that is off as Felhom runs them; the row reminds us never to enable EE features or the -enterprise meilisearch image.", "cost": "Keeping a standing open row; the real guard would be a line in the licence table / REUSE.md read on each major.", "if_never": "Nothing changes unless someone turns on an EE feature; the rule survives in the licence audit.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-794", "sev": "P4", "category": "Business & legal", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog@917a779 seven redis 7 images: dawarich/docker-compose.yml:164 `image: redis:7.4-alpine`, docmost:86, immich:123, outline:88, nextcloud:103, paperless-ngx:125, romm:136 `image: redis:7-alpine`. Moving each needs a harness-proven ladder step (7 apps).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-796", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "app-catalog@917a779 templates/metube/.felhom.yml:19 `family_gate: true` with no family_gate_except (by design, per row).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "MeTube's browser/phone 'send to MeTube' helpers cannot pass the family gate; households paste links in the page.", "cost": "A per-member token the gate accepts on /add only - a new design.", "if_never": "Households use the page; the helpers stay unusable. Reopen if a household asks.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-797", "sev": "P4", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "app-catalog@917a779 scripts/check-family-gate.py:126 `print(\"family-gate: rule 3 NOT CHECKED — no felhom.eu sibling with a baked golden at %s\" % sibling)` and :141 ' — rule 3 (golden >= 0.287.0) NOT CHECKED here' - the gap is stated, and the pre-push hook (with sibling) checks it.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "CI's single-repo clone cannot check rule 3 of the family-gate gate; it says NOT CHECKED, and the pre-push hook checks it.", "cost": "Giving catalog CI a felhom.eu sibling checkout (credentials, workflow change).", "if_never": "A push that bypasses the hook (--no-verify is forbidden) could skip rule 3; CI stays honest about it.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-798", "sev": "P4", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog@917a779 templates/grimmory/docker-compose.yml:29 ` - SWAGGER_ENABLED=false` still present.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": "templates/grimmory/docker-compose.yml", "change": "Remove the dead SWAGGER_ENABLED line (or rename to API_DOCS_ENABLED=false, which v3.5.0 reads and defaults to false). No image: line moves.", "test": "catalog_gates.py --fast green; grep shows no SWAGGER_ENABLED. Note: a compose change syncs to boxes running grimmory and recreates the container, so the row's 'on the next Grimmory step' timing is reasonable.", "minutes": 10}, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-799", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog@917a779 scripts/upgrade_fixtures_box.py:2183 `data=json.dumps({\"url\": self.URL, \"quality\": \"best\", \"format\": \"any\", \"auto_start\": True}), method=\"POST\")` - no download_type.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": "scripts/upgrade_fixtures_box.py (class MeTube.seed)", "change": "Add \"download_type\": \"video\" to the POST /add body.", "test": "Python syntax/import check of the fixture module; real proof rides the next MeTube bench/box walk (POST /add 200).", "minutes": 10}, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-804", "sev": "P4", "category": "Apps & catalog", "group": "NOT-WORTH-IT", "evidence": "app-catalog@917a779 templates/plant-it/docker-compose.yml:12 `image: msdeluise/plant-it:0.10.0`; templates/plant-it/.felhom.yml:23-26 \"The compose below is deliberately LEFT AS-IS (it pins `msdeluise/plant-it:0.10.0`, a repository ... It is not worth fixing\" and `lifecycle: abandoned`.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "plant-it's image repository does not exist; the template is already abandoned and not installable, and no box runs it.", "cost": "An operator choice to hide the template entirely (small catalog edit).", "if_never": "Nothing; the template is already documented as not installable on purpose.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-805", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog-felhom.eu scripts/check-volume-persistence.py:342 'if m[\"class\"] == \"named-declared\" and m.get(\"files\", 0) == 0:' — only named volumes judged empty; binds not. Last change 917a779 (R-788). The rule change itself is small, but it flips Grimmory/komga/paperless-ngx/radarr/sonarr to UNDETERMINED and needs a live re-sweep to regenerate the verdict tables; the row also asks for a decision.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-806", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "app-catalog-felhom.eu scripts/check-volume-persistence.py:586 'code = _sh(a + [f\"http://{ip}:{port}{path}\"], timeout=40)' — plain http always; scheme reading exists only in scripts/upgrade_boxport.py:33 LB_SCHEME_RE and scripts/upgrade-test.py:407. The gramps-web no-answer half is not checkable from source.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["scripts/check-volume-persistence.py", "scripts/test_check_volume_persistence.py"], "change": "Make routed_ports return the traefik loadbalancer.server.scheme (reuse upgrade_boxport LB_SCHEME_RE) and have the GET exercise use that scheme with curl -k for https. The gramps-web :5000 non-answer stays a separate live look (narrow the row to it).", "test": "Unit test: a compose with scheme=https label yields an https:// URL in the built curl argv; decoy without the label yields http://.", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-807", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Per-app upload seeds for 13 apps (claper, crafty-controller, dawarich, docmost, gramps-web, immich, outline, sparkyfitness, tandoor, vikunja, wger, wishlist, zipline) + plex/wanderer; each needs a live fixture run. Gate rule at scripts/check-volume-persistence.py:342 still makes empty declared volumes UNDETERMINED (917a779).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-814", "sev": "P4", "category": "Hub & operator", "group": "UNCHECKED", "evidence": "Live Hetzner console state (box 611421 status); not visible in source. Operator action only.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-815", "sev": "P4", "category": "Backup & restore", "group": "UNCHECKED", "evidence": "First GC completion on felhom-offsite is PBS server-side live state; grep of documentation found no GC completion record (DIAG-backup-missed-2026-07-26.md:43 'prune/GC history NOT COLLECTED').", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-816", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Needs a live exercise of six failure classes on a scratch guest; no source change can close it.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-817", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu documentation/architecture/09-update-architecture.md:619 '56. **A box keeps the controller image it runs and the one before it** (the self-update's roll-back target)'. Source disagrees: felhom-agent internal/localapi/controllerswap.go:240 'st.Previous, st.Current = prev, prev' (prev = image running when the swap began); felhom-controller controller/internal/stacks/controller_image_retention.go:21-22 '...the self-update's own roll-back needs only the running image... \"The previous\" is kept for a hand roll-back.' The decision text is the wrong one.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["documentation/architecture/09-update-architecture.md"], "change": "Add a dated correction under decision 56: the swap rolls back to the image running when the swap began (controllerswap.go Swap/rollback); the kept previous image is for a hand roll-back. Do not rewrite the ruling itself.", "test": "Doc only: grep the corrected line; repo_gates.py --fast passes.", "minutes": 15}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-818", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu hub/CHANGELOG.md:759 '## v0.109.0 — the Backup card told every operator that every customer had no backups (2026-08-30, R-331)' and :766 '(R-330 is a nightly false alarm ...; R-331 is a hub display ...)'; OPEN-ITEMS.md:268/274 have R-330/R-331 as Disk health Phase 2/3. Line numbers moved from the row's 526-538 to 759-771. No correction line present.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/CHANGELOG.md"], "change": "Add a dated correction note under the v0.109.0 entry (and the hub CHANGELOG mentions at :695/:701) naming the real closed rows from CLOSED-ITEMS.md. Also present in felhom-controller/CHANGELOG.md:3107 and :3234 (R-330/R-331 for v0.224.0/v0.225.0) — a second repo; either note it there too or narrow the row.", "test": "Doc only: grep the correction; gates pass.", "minutes": 25}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-819", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "Ran python3 scripts/check_stands.py (read-only): still convicts e.g. 'fail.stolen-machine: register id R-281 is not in OPEN-ITEMS.md', 'fail.customer-self-restore: register id R-356 ...'. grep 'stands' in scripts/repo_gates.py and .gitea/workflows/gates.yml: no hit. Last commit on the script 6088afcb.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/check_stands.py", "scripts/repo_gates.py", "where-felhom-stands.yaml (or its source)"], "change": "Let rule 3 accept an id found in CLOSED-ITEMS.md (and check the stand's status agrees), fix the R-273/R-356 dangling ids, register the gate in repo_gates.py with a decoy.", "test": "Run check_stands.py green; a decoy stand citing a non-existent id must fail; test_repo_gates.py passes.", "minutes": 60}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-832", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Deferred roadmap item: a third-location copy of ep0 is money + operator decision (decision 71). Nothing in source to change.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-844", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Needs a household timeline on the controller, which does not exist (row: 'when the box gets a household timeline'). A new surface, not a fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-855", "sev": "P4", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu hub/cmd/hub/main.go:450 'logger.Printf(\"[INFO] osupdates: the Docker engine set is approved only by the operator, after %d healthy ring-0 night(s)\", osSvc.DockerNights)' prints the raw value; internal/osupdates/service.go:169 'negative means none'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/cmd/hub/main.go", "hub/internal/osupdates/service.go"], "change": "Add a small helper (e.g. osSvc.DockerNightsEffective()) mapping negative→0 and 0→2, and print that in the start log.", "test": "Unit test of the helper for -1, 0, 3; red-proof by printing the raw value.", "minutes": 20}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-856", "sev": "P4", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Row is marked an operator design question (crash-restart suppression for app mails); not a defect yet.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-857", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu scripts/golden_currency_gate.py:146 'EVIDENCE_RE = re.compile(r\"^golden-(\\d+)\\.(\\d+)\\.(\\d+)-\\d{4}-\\d{2}-\\d{2}$\")' and :237-238 'found.sort() / return found[-1]' — two dirs with the same version tie-break by name, not by bake time; a '-rebake' suffix dir (documentation/tests/golden-0.292.0-2026-10-04-rebake/) does not match the regex at all, so the re-bake is never read.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/golden_currency_gate.py", "scripts/test_golden_currency_gate.py", "documentation/runbooks/RUNBOOK-manual-build.md (re-vouch line)"], "change": "Have newest_baked accept an optional suffix after the date and, for equal versions, prefer the newest bake-log timestamp (or refuse two dirs for one version); add 're-vouch at once after a same-version re-bake' to the runbook.", "test": "Test with two tmp dirs of one version holding different GOLDEN_SHA256 lines: the newer sha must be reported; red-proof against today's code.", "minutes": 45}, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-878", "sev": "P4", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Next action is 'measure a large volume first' — a live measurement; the fix direction is a behaviour change to the catch-up.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-881", "sev": "P4", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu scripts/felhom-host-install.sh: grep 'felhom-priv-apply' → no hit anywhere in the installer (uninstall does not remove it); :1679 '+ guest-hook snippet under /var/lib/vz/snippets/ (agent-installed at runtime)' still says runtime-installed.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh", "CHANGELOG"], "change": "Add an rm -f of /usr/local/sbin/felhom-priv-apply to the uninstall step (tolerate-absent, like felhom-pbs-apply at :1161) and fix the :1679 comment; ships at the next installer tag.", "test": "bash -n; --uninstall --dry-run output lists the removal; installer gates pass.", "minutes": 25}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-884", "sev": "P4", "category": "Monitoring & notifications", "group": "UNCHECKED", "evidence": "Live ArgoCD diff on DooPlex (forbidden to touch here). Related: homelab-manifests mon-system/monitoring.yaml:399-426 is the same prometheus Deployment R-211 concerns.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-885", "sev": "P4", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "On main (b018ca90) scripts/repo_gates.py runs no test_*.py suite. NOTE: the felhom.eu working tree holds UNCOMMITTED work for exactly this row by another session: '?? scripts/script_tests_gate.py' (docstring 'every Python test suite under scripts/ runs on every push (R-885)'), '?? scripts/test_script_tests_gate.py', ' M scripts/repo_gates.py' (registers 'script-tests'). Not fixed until pushed.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/script_tests_gate.py", "scripts/test_script_tests_gate.py", "scripts/repo_gates.py"], "change": "Finish and push the in-progress script_tests_gate.py (walks scripts/ for test_*.py, exit-code verdict, nesting guard) registered in repo_gates.py; coordinate with the session that owns the dirty tree.", "test": "test_script_tests_gate.py decoys (a failing suite, no suite found); one CI run green.", "minutes": 30}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-30", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Design change (presence from the Dir-2 long-poll instead of the report clock), size M; no commit with R-30 after aa9c08f0 (filing).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-31", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu hub/internal/web/configs.go:1594 'd, err := s.offsite.ProvisionOffsite(ctx, cfg.CustomerID, in)' still in-request; :1584 detaches from the request context (mid-cancel fixed) but no async/status card.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-35", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller controller/internal/report/config_refresh.go:65 'config-refresh: applied config_version=%d — self-restarting to load it'; sessions are in-memory only: controller/internal/web/auth.go:259-260 's.sessions[token] = &session{'. Hot-apply or persisted sessions is a design change with security weight.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-49", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog-felhom.eu templates/immich/.felhom.yml:28-34 backup block has no cache/volume exclusion; row itself says a capture-set exclusion needs its own ruling (data-loss-shaped).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-50b", "sev": "P3", "category": "Box system & updates", "group": "FIXED-BY-LATER-WORK", "evidence": "Claim 'fetched via fetch_raw from raw/branch/main — no tag, no pin' no longer true: bee68484 (installer v1.23.0, R-110/R-183) pinned fetch_raw to the vouched agent tag — felhom.eu scripts/felhom-host-install.sh:533 '\"$GITEA_BASE/$GITEA_OWNER/$AGENT_REPO/raw/tag/v$ART_AGENT_VER/$path\" \\' (leg b). Leg (c)-like signed delivery: felhom-agent c9fa2e7 (R-840 config bundle) and configs/test_felhom_config_bundle.py:264 covers /usr/local/sbin/felhom-pbs-apply. Residual worth one line if kept: the 0440 sudoers drift visibility note.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-78", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "An owed operator decision + spike (local_api authority); not a code defect.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-79", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller controller/internal/monitor/healthcheck.go:100 'fmt.Sprintf(\"SSD disk usage critical: %.0f%%\"', :213 'Protected container not running: %s'; rendered raw at controller/internal/web/alerts.go:241 'Message: issue, // ON THE WIRE ... not ours to translate; slice 3'. Whole-surface, seam needs a spike.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-118", "sev": "P3", "category": "Storage & devices", "group": "STILL-TRUE-SMALL", "evidence": "felhom-agent internal/localapi/disks.go:401 'if total, used, okc := statfsCapacity(d.MountPath); okc {' — no device-presence guard on the union path; devicePresent exists (disks.go:985) and is used just above (:381) for BoundUnderParent.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-agent", "files": ["internal/localapi/disks.go", "internal/localapi/disks_device_presence_test.go"], "change": "Only statfs the mount when s.devicePresent(d.MountPath) is true (else leave capacity zero/unknown); put statfsCapacity behind a seam var for the test.", "test": "Test: absent device + statfs seam returning root's numbers → row has no capacity; present device → capacity set; red-proof by removing the guard. Ships in the next agent release.", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-121", "sev": "P3", "category": "Box system & updates", "group": "FIXED-BY-LATER-WORK", "evidence": "3d7a2761 (hub v0.135.0, R-530/R-604 'boxes left behind listed and alarmed'): felhom.eu hub/internal/osupdates/service.go:79 'EventAgentBehind = \"agent_behind\" // warning, operator' with :180 'AgentBehindAfter: a box runs an agent older than the vouched one this long → an operator alarm' (7 d window, the staleness window the row asked for).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-126", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller controller/internal/web/handler_export.go:377-386 storageDriveList() appends every s.settings.GetStoragePaths() entry with no IsNetwork() filter; the predicate exists at controller/internal/settings/settings.go:618 'func (p StoragePath) IsNetwork() bool { return p.Kind == StorageKindNetwork }'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/handler_export.go", "controller/internal/web/handler_export_test.go"], "change": "Skip IsNetwork() paths in the export-destination list and refuse them in isValidDrivePath for the export POST (keep scanning for import if wanted, via a separate list).", "test": "Handler test with one drive + one network path: destination list has only the drive; export POST to the NAS path is refused; red-proof without the filter. Needs a controller release.", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-127", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Leg (a) still true: grep 'data_key: true' in app-catalog-felhom.eu templates → only adventurelog, dawarich, homebox, papra, sparkyfitness; n8n N8N_ENCRYPTION_KEY, wanderer POCKETBASE_ENCRYPTION_KEY, calcom CALENDSO_ENCRYPTION_KEY, bookstack APP_KEY unflagged (templates/n8n/.felhom.yml:38 etc.). Leg (b) (regenerated DB password vs restored PGDATA) needs a design choice. Leg (a) alone is a ~45-min catalog change + label/flag agreement gate and could be split out.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-130", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu scripts/felhom-host-install.sh:348 'HARD_MIN_LVM_GIB=120 # a useful appliance won't fit below this on local-lvm' and :1760 '... || log_warn \"local-lvm free ~${free_gib} GiB < hard min ${HARD_MIN_LVM_GIB} GiB\"' — warns only.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh"], "change": "Take the cheap honest branch: rename to RECOMMENDED_MIN_LVM_GIB and reword the warning to 'below the recommended …' (making it refuse would change install behaviour and needs a ruling).", "test": "bash -n; installer gates; grep shows no 'hard min' left. Ships at the next installer tag.", "minutes": 20}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-132", "sev": "P3", "category": "Security & access", "group": "UNCHECKED", "evidence": "Whether HUB_PW was rotated is out-of-band operator state; not visible in source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-136", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu hub/internal/web/server.go:857 'Name: \"hub_session\",' and readers at server.go:806, :896, :918 and apps.go:351.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["hub/internal/web/server.go", "hub/internal/web/apps.go", "a server test"], "change": "Introduce a const sessionCookieName = \"__Host-hub_session\" and use it at all five sites (Path=/, Secure, no Domain already hold). Every operator logs in once more; plain-HTTP browser access stops (Basic auth unaffected).", "test": "Test: login response sets __Host-hub_session with Secure+Path=/ and no Domain; a request with the old name is not a session.", "minutes": 30}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-137", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller controller/internal/cloudflare/waf.go:18 'globalRuleDesc = \"[felhom-geo] Global\"', :21 'appRuleDescPrefix = \"[felhom-geo] app:\"' — still not namespaced. Two-repo M change.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-138", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller controller/internal/infra/infra.go:157-159 writes CF_DNS_API_TOKEN when d.CFAPIToken != \"\"; no shared-zone guard in hub (grep shared.zone: none). Needs a policy decision first; no shared zone exists today.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-179", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu scripts/felhom-host-install.sh uninstall section :1121-1157 handles felhom-shared-parent and umounts under /mnt/felhom-drives, but grep 'x2ddrives|automount' → no hit: the per-share mnt-felhom\\x2ddrives-*.mount/.automount units are never stopped or removed.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh"], "change": "In uninstall step 4c, before the umount loop: stop+disable every mnt-felhom\\x2ddrives-*.automount/.mount unit, rm their files from /etc/systemd/system, daemon-reload (tolerate-absent; still never umount -l/-f).", "test": "bash -n; --uninstall --dry-run on a fixture listing; a demo-box uninstall run is the real check at the next installer tag.", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-180", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu scripts/felhom-host-install.sh:1800 'if pvesm status --storage \"$ARCHIVE_STORAGE\" ...' checks existence only; PVE_STORAGES=(local local-lvm felhom-pbs) at :322; no ARCHIVE_STORAGE ∈ PVE_STORAGES assertion.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh"], "change": "In the same pre-flight block, die (byo) / die (appliance) if ARCHIVE_STORAGE is not in PVE_STORAGES, with a message naming --acl-storages.", "test": "bash -n; a dry-run with --archive-storage felhom-backup must die in pre-flight; with local passes.", "minutes": 25}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-190", "sev": "P3", "category": "Box system & updates", "group": "NOT-WORTH-IT", "evidence": "Mitigation still in source: felhom-agent cmd/felhom-agent/main.go:656 'store-grant: GRANT WAS MISSING AND HAS BEEN SELF-REPAIRED — investigate the loss (R-190)'. Mechanism unexplained since 2026-08-03.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A storage permission vanished once on demo-felhom in August and nobody knows why. Since agent 0.124.1 the agent puts it back by itself and mails the operator when it happens.", "cost": "Finding the cause needs live experiments with guest rebuilds and permission caches on a Proxmox host — hours, maybe days, with no guaranteed answer.", "if_never": "The box repairs the permission within one check cycle and the operator gets one e-mail each time; a recurrence would be visible and would reopen the question.", "pick": "close-as-accepted"}, "minutes_spent": 4}
|
||||
{"id": "R-200", "sev": "P3", "category": "Backup & restore", "group": "FIXED-BY-LATER-WORK", "evidence": "The remaining half (customer-facing recovery-code form: yell → R form → preview) shipped as the recovery screen: felhom-controller 636c51e 'R-193: the recovery screen — unlocking, and only unlocking (v0.200.0)'; controller/internal/web/templates/recovery.html:82 '<form id=\"unlock-form\" method=\"POST\" action=\"/recovery/unlock\" autocomplete=\"off\">', routed at internal/web/server.go:602. Plumbing half was 1b1366b (v0.196.0). The 64-hex inject-password route (server.go:783) stays a deliberate DR fallback with no form.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-211", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "homelab-manifests (/home/kisfenyo/git/homelab-manifests @87dfc29) mon-system/monitoring.yaml:420 'image: prom/prometheus:v3.15.0', :426 '--web.enable-lifecycle'; grep 'reload|checksum/config' → none. The manifest edit is small, but it rolls the production Prometheus on DooPlex (operator territory) and the same Deployment is OutOfSync per R-884 — do the two together.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-231", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Owner operator; DooPlex /opt/backup/scripts remains host state. Partially touched by cea8502f (scripts/hub-db-backup versioned in felhom.eu, cites R-231) but that covers only the hub-DB push, not /opt/backup/scripts or the same-disk/no-off-site facts.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-235", "sev": "P3", "category": "Install & onboarding", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu c033b3b6 'ISO 1.28.0 source: the console stops showing the pairing code once bound (R-535)'. scripts/iso/felhom-bootstrap.sh:538: `print_bound_banner # R-535: replace the pairing code on the console with the truth`. Same defect already CLOSED twice in CLOSED-ITEMS.md as R-535 (line 232) and R-214 (line 200, 'proven on a fresh install').", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-240", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/backup/offbox.go:1168: `warns = append(warns, \"Sikeres — nincs mentésre jelölt alkalmazás\")`; web/handlers.go:1068-1069 returns lastWarning verbatim when toggledCount < 1, so the customer still sees 'Sikeres'. Run now also records warnKind=OffboxWarnNoAppsSelected (R-553), so the page no longer depends on the sentence for new runs.", "dup_of": null, "unique_facts": "handlers.go:1064-1066 comment: R-557 must not TRANSLATE the producer until R-570 closes (legacy kind=='' fallback matches the lowercase substring 'nincs mentésre jelölt alkalmazás'). A Hungarian rewording that keeps that lowercase substring stays safe for legacy boxes.", "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/backup/offbox.go", "controller/internal/backup/offbox_test.go", "controller/internal/backup/offsite_diag_test.go", "controller/internal/backup/r553_offsite_quota_test.go", "controller/internal/web/r553_stale_note_test.go"], "change": "Replace the producer string at offbox.go:1168 with wording that drops 'Sikeres' but keeps the lowercase marker substring, e.g. 'Ez a futás semmit nem mentett: nincs mentésre jelölt alkalmazás'; update the tests that pin the literal.", "test": "go test ./internal/backup ./internal/web -run 'Offbox|R553|Diag' (unit, no docker); red-proof: a test asserting the warning does not start with 'Sikeres' fails on old code.", "minutes": 45}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-242", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/scripts/golden_currency_gate.py:23-24: 'It does **NOT** check that the golden was **VOUCHED**, because the vouched version lives ONLY in the hub's `hub_settings` table'; :40 'That vouch half is STILL open after 2026-09-13'. Last gate commits 5ef0f52b/ae59c31a did not add a vouch check.", "dup_of": null, "unique_facts": "Only the vouch half remains; it needs a hub-reading check (design: shape (c) hub-side checker) and cannot be a --fast gate. Operator ruling 2026-09-13 (goldens weekly + waiver file) reduces the urgency; candidate for NOT-WORTH-IT if the operator accepts.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-244", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub: no cascade leg touches app_log_issues — only writers are internal/store/telemetry.go:176-209 (upsert), :537 `DELETE FROM app_log_issues WHERE last_seen < ?`, :556/:575 operator deletes by app/id. cmd/hub/main.go:1004: `if n, err := s.PruneStaleIssues(time.Now().Add(-30 * 24 * time.Hour))` (since a757bee0, hub v0.4.0).", "dup_of": null, "unique_facts": "Not in the row: a 30-day stale-issue prune has existed since hub v0.4.0, so ORPHAN rows (only torn-down customers) age out on their own once not seen for 30 days — the 'accumulates one venue at a time' claim holds only for the SHARED rows, which keep a deleted customer's id in affected_customers while a live customer keeps reporting. The remaining fix is JSON de-referencing in the cascade + red-proof.", "small_fix": null, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-250", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub/internal/offsite/offsite.go:71-72: `var defaultScanBackoff = []time.Duration{2 * time.Second, 4 * time.Second, 8 * time.Second, 16 * time.Second, 30 * time.Second}`; scanner.go:40 dials plain \"tcp\" (no A-record preference); no R-250 commit.", "dup_of": null, "unique_facts": "Not in the row: cmd/hub/main.go:617 `WriteTimeout: 60 * time.Second` — the ~60 s ladder already meets the server write deadline, so lengthening the ladder inside the POST needs a per-request deadline lift (internal/api/wait.go:58 shows the pattern) or async provisioning; that is why it is not a one-line change.", "small_fix": null, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-251", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/backup/offbox_inventory.go:102-108: loops `for _, tag := range sn.Tags` and adds every non-empty tag to `newest` — no filter for 'felhom-offbox'. The marker filter exists only on the restore list (internal/web/offsite_restore_list.go:37 `const offboxMarkerTag = \"felhom-offbox\"`), not on the recovery inventory (web/recovery_handlers.go:175 `data[\"InvApps\"] = inv.Apps`).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/backup/offbox_inventory.go", "controller/internal/backup/offbox_inventory_test.go"], "change": "In offsiteNewestPerTag skip the 'felhom-offbox' marker tag (move the constant into package backup and reuse it from web); consider whether '_shares' should render as a named row or be skipped.", "test": "Extend inventoryFixture (offbox_inventory_test.go:22) so snapshots carry tags [felhom-offbox, <app>]; assert OffsiteInventoryList returns only the app rows and one size call per app; red-proof on current code (marker row appears).", "minutes": 40}, "not_worth": null, "minutes_spent": 9}
|
||||
{"id": "R-255", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller: still no page-wide runtime secret-sentinel test — `grep -l sentinel internal/web/*_test.go` hits only edge_safe_status_test.go, i18n_parity_test.go, recovery_test.go; controller/scripts/secret_in_markup_gate.py remains the only all-template net. No commit cites R-255.", "dup_of": null, "unique_facts": "Not in the row: the i18n parity test (internal/web/i18n_parity_test.go, TestI18nParity:516, 133 fixtures in testdata/i18n_parity) now builds per-page render data for most pages — the per-page fixture scaffolding this row priced as the main cost largely exists; a secret-sentinel pass could reuse it, making this cheaper than estimated (still > 1 h).", "small_fix": null, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-257", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/i18n/locales/hu.json:1409: `\"flash.offbox.not_orphaned\": \"Az offsite tároló nincs elárvult állapotban.\"`, used at internal/web/offbox_handlers.go:317 (row cited :270); also hu.json:1216 err.backup.az_offsite_tarolo_nincs_elarvult_allapotban from backup/offbox.go:385.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/i18n/locales/hu.json", "controller/internal/i18n/locales/en.json"], "change": "Rewrite flash.offbox.not_orphaned (hu + en) to say what the customer tried, that it does not apply now, and where to look, without 'offsite'/'elárvult', e.g. 'A távoli mentés rendben van, nincs mit félretenni. Ha gondod van vele, írj nekünk.'; wording sign-off from the operator (owner Viktor).", "test": "Run the i18n parity/key tests (go test ./internal/i18n ./internal/web -run I18n) and an ASCII grep for 'elarvult' in the flash value as negative control.", "minutes": 25}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-262", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/hub/internal/api/handler.go:727-729: '// hostBackup / hostRestoreTest mirror the agent's hub.Backup / hub.RestoreTest wire // contract field-for-field'; struct hostRestoreTest (handler.go:746-761) has no mount_parity/mount_inventory, while felhom-agent/internal/hub/report.go:492-493 `MountParity string json:\"mount_parity,omitempty\"` / `MountInventory []string`. grep finds no mount_parity anywhere in hub Go code.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/internal/api/handler.go", "hub/internal/api/host_test.go"], "change": "One-repo fix: narrow the comment to say hostBackup is field-for-field and hostRestoreTest is a deliberate SUBSET (lists mount_parity/mount_inventory as not modelled), and add a hub test that names the two agent fields as known-unmodelled so a future addition must edit it. Adding the fields + fixture is a two-repo change (byte-identical golden) and stays a separate choice.", "test": "go test ./internal/api -run HostReport (hub, no docker).", "minutes": 30}, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-269", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom-agent/internal/localapi/tokenstore.go:173-176: `if vmid, ok := s.byHash[want]; ok { if subtle.ConstantTimeCompare(...) == 1 { return vmid, true } }` — a superseded token's hash is still a direct hit; reload happens only on a miss (:180). Last change to the file f31a76f (v0.63.0, the reload-on-miss itself); no R-269 commit.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-agent", "files": ["internal/localapi/tokenstore.go", "internal/localapi/tokenstore_test.go"], "change": "On a map hit, also stat the store and reload when its size differs from loadedSize before answering (one cheap stat per auth), so a rotated-out token is rejected without depending on an unrelated miss.", "test": "Add TestTokenStore_RotatedOutTokenRejectedFirst: mint A, re-mint B from a second TokenStore instance on the same file, look up A FIRST on the long-lived instance -> want false. Red on current code (row says red-proved).", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-270", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/internal/bootstrap/bootstrap.go:255: `if cfg == nil || cfg.LocalAPI.Endpoint != \"\" {` (fill-only, never refreshes); DetectEndpointDrift (bootstrap.go:369-399) compares only the endpoint. No R-270 commit.", "dup_of": null, "unique_facts": "Blocked in substance on R-78 (OPEN, OPEN-ITEMS.md:323) — the operator's authority ruling (auto-reconcile vs detect-only) decides which fix is right. The rotation recipe itself is documented in memory 'local-API token rotation needs TWO files + TWO restarts'.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-271", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/internal/channelhealth/checker.go:152: `if prev != \"\" && prev != \"up\" {` (unseeded->up is silent); checker.go:87 alert text still says '(re-bootstrap)'. No R-271 commit.", "dup_of": null, "unique_facts": "Fix needs a persisted last state or a hub-seeded state (a new mechanism), so not an under-an-hour change.", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-274", "sev": "P3", "category": "Box system & updates", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu eb600872 'R-297: installer compares a local golden against the manifest before using it'. scripts/felhom-host-install.sh:3074: `if golden_local_matches_manifest \"$GOLDEN_VOLID\"; then` (digest vs ART_GOLDEN_SHA, else baked marker vs ART_GOLDEN_VER; otherwise ignores the local golden and fetches, or dies if the operator named it, :3080). Line numbers differ from the triage note (2855-2905).", "dup_of": null, "unique_facts": "The BYO-disclosure half ('say what the install REUSES') was not checked; if wanted, it is a separate small wording row.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-275", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/felhom-host-install.sh:1059: `for _cfgbak in \"${agent_cfg}\".bak*; do [[ -e \"$_cfgbak\" ]] && run rm -f \"$_cfgbak\"; done` — glob still misses agent.json.campaign8-before, .campaign9-prev, .pre-e-target-move, .pre-prunegate.bak; :814 still claims 'config (+ its .bak backups)'. No R-275 commit.", "dup_of": null, "unique_facts": "The uid-reuse half (new service account inheriting uid 999) is not covered by the small fix; purging by directory makes it moot for /etc/felhom-agent.", "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh", "scripts/hostinstall_gates.py"], "change": "Purge every sibling `\"${agent_cfg}\".*` (and then rmdir/rm the config dir) instead of `.bak*`; fix the WIPED line wording.", "test": "Add a hostinstall_gates.py check (or a bash harness under a temp dir with FELHOM paths) that creates agent.json.campaign8-before and agent.json.pre-prunegate.bak and asserts run_uninstall --dry-run lists both; red on current glob.", "minutes": 40}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-276", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/felhom-host-install.sh: no reference to wg-felhom/wg-quick anywhere (grep 'wg-quick\\|wg-felhom' empty); _uninstall_statement (:807-845) lists neither in WIPED nor KEPT. The tunnel is agent-managed: felhom-agent/internal/wgtunnel/manager.go:31 `confDest = \"/etc/wireguard/wg-felhom.conf\"`, :35 `unit = \"wg-quick@wg-felhom\"`.", "dup_of": null, "unique_facts": "Hub-side peer deregistration (hub/internal/store/wg_operator.go) is a second repo; the small fix covers the host side and names the hub peer under KEPT.", "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh", "scripts/hostinstall_gates.py"], "change": "In run_uninstall (full scope): stop+disable wg-quick@wg-felhom and remove /etc/wireguard/wg-felhom.conf via run(); add it to WIPED, and add a KEPT line 'the hub-side WireGuard peer registration — remove it in the operator UI'.", "test": "Static gate in hostinstall_gates.py asserting the uninstall path names wg-quick@wg-felhom and the statement mentions it; --dry-run output check.", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-277", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "Part (a) FIXED by felhom.eu f5c9411e 'R-331 (hub half): the Backup card reads `offsite`, not the dead `backup` fields (v0.109.0)' (hub/internal/web/backup_card.go uses fmtBytesAuto :135-142). Part (b) still true: hub/internal/web/offsite_box.go:54 `return fmt.Sprintf(\"%.1f GB\", float64(b)/float64(int64(1)<<30))` — a 162 KB repo renders 0.0 GB.", "dup_of": null, "unique_facts": "Part (c) (stale offsite_delivery_stuck event read as current state) was NOT verified; if it is still wanted it should become its own row.", "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/internal/web/offsite_box.go", "hub/internal/web/offsite_box_test.go"], "change": "For UsageStr use fmtBytesAuto when the usage is below 1 GB (keep GB for the quota and the bar), so a non-empty repo never renders as 0.0 GB.", "test": "Unit test: 162*1024 bytes -> not '0.0 GB'; 0 bytes and >1 GB unchanged (go test ./internal/web -run Offsite).", "minutes": 25}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-282", "sev": "P3", "category": "Install & onboarding", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu 4d6ec7c 'hub v0.104.0: ... the hub half of the naming (R-295)' + controller v0.211.0 (R-295 CLOSED, CLOSED-ITEMS.md:207) + R-323 hub v0.105.0. hub/internal/notify/templates.go:204: '// R-295, HUB HALF (2026-08-13). ONE NAME PER SECRET, and it is „Beállító kód\".'; hub/internal/claim/engine.go:51 `EmailReenroll EmailKind = \"reenroll\"` (mail names the setup page a rebuilt box shows).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-283", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub/internal/claim/engine.go:7: '// engine: a rotation bumps the generation (single active code) and NEVER clears claimed_at.'; ReissueForReenroll (engine.go:195-212) rotates and mails but leaves the claim set — the hub still shows the customer as claimed after a guest rebuild. No R-283 commit.", "dup_of": null, "unique_facts": "The mail half is now right (EmailReenroll names the setup page, R-295 hub), so the R-282 consequence is gone; what remains is hub/box claim-state reconciliation (design).", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-298", "sev": "P3", "category": "Storage & devices", "group": "UNCHECKED", "evidence": "The template gate is still there: felhom-controller/controller/internal/web/templates/storage.html:364 `if(d.role==='user-data'){` else protected (:368). BUT the agent no longer reclassifies a backup-target drive: felhom-agent/internal/localapi/disks.go:1230-1236 ('WHY THIS IS NOT A ROLE RECLASSIFICATION ... the drive that now holds the whole-guest archives is ALSO the enrolled user-data drive') and disks.go:219/227 report Role from RoleForStorage plus a separate BackupTarget flag (958e54f, agent v0.112.0). storage/role.go:176-186 makes a local-dir with its own non-system device 'user-data'.", "dup_of": null, "unique_facts": "From source, a drive that is both user-data and the backup target should arrive as role 'user-data' and therefore be registrable; the row's 2026-08-10 observation contradicts that. Only a live /api/disks read on a box with that layout settles it.", "small_fix": null, "not_worth": null, "minutes_spent": 10}
|
||||
{"id": "R-306", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom.eu/scripts/felhom-host-install.sh:418: `$DRY_RUN && return 0` (only DRY_RUN short-circuits _state_put); :1851/:1854 `_state_put dnsmasq_preexisting yes|no` run unguarded in preflight, while :226 says '--preflight-only: ... no state writes' and :1956-1961 guards only customer_id/mode. Blockers R-300 and R-305 are CLOSED.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/felhom-host-install.sh", "scripts/hostinstall_gates.py"], "change": "Make _state_put (and _state_mark) return 0 when PREFLIGHT_ONLY is true, so a preflight-only run writes nothing; the real run's preflight records ownership again.", "test": "Bash harness with STATE_DIR in a temp dir: run the preflight state block with PREFLIGHT_ONLY=true -> no state.json; red on current code. Or a static gate asserting _state_put checks PREFLIGHT_ONLY.", "minutes": 30}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-314", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller: StopAbandon is called only from cmd/controller/main.go:244 (CLI) — `grep -rn StopAbandon` shows internal/backup/offbox_abandon.go:374 and that CLI call, no web handler.", "dup_of": null, "unique_facts": "An 'operator-authenticated POST' on the controller needs an operator auth path the controller does not have today (or a hub-relayed command) — a new mechanism.", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-317", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-SMALL", "evidence": "felhom-agent/internal/lanresolver/lanresolver.go:107: `if _, err := os.Stat(\"/usr/sbin/dnsmasq\"); err != nil { // metadata read, no privilege needed` — still probes the dnsmasq-base file. No R-317 commit.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-agent", "files": ["internal/lanresolver/lanresolver.go", "internal/lanresolver/lanresolver_test.go"], "change": "Probe the dnsmasq unit (e.g. /usr/lib/systemd/system/dnsmasq.service or /lib/systemd/system/dnsmasq.service) behind a small stat seam instead of /usr/sbin/dnsmasq.", "test": "Seam test: binary present + unit absent -> install is attempted; unit present -> skipped. Red on current code.", "minutes": 40}, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-330", "sev": "P3", "category": "Storage & devices", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-agent/internal/hub/report.go:406-425 SmartSummary carries reallocated/pending/offline_uncorrectable + NVMe set only — no 187/188/199 fields. Wire change across agent + hub (+ controller), declared M.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-332", "sev": "P3", "category": "Storage & devices", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Closing condition is live-only (a real degrading disk or an injection through agent /disks -> controller -> hub). Last related commits ea16a21b/2fa1efc narrowed the restart half only; no commit records a live Hiba-from-counters verdict.", "dup_of": null, "unique_facts": "Needs an injection harness (new mechanism) or a real failing disk; cannot close from source.", "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-333", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "(b) felhom-agent/internal/storage/hostops.go:374: `out, stderr, err := h.runner.Run(ctx, h.bins.Smartctl, \"-a\", \"-j\", device)` — no -n standby. (a) still an operator decision (Viktor decides).", "dup_of": null, "unique_facts": "Not in the row: (b) is not a one-line agent change — configs/felhom-agent.sudoers:34 pins the exact argv `/usr/sbin/smartctl ^-a -j /dev/(...)$`, so adding `-n standby` also needs a sudoers change delivered by the signed bundle.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-338", "sev": "P3", "category": "Security & access", "group": "UNCHECKED", "evidence": "felhom.eu/documentation/operations/nodes.md:86-88 now says demo-hp 'Agent config shape (R-50 island): local_api on 169.254.253.1:8443/vmbr9, guest eth1 169.254.253.2/30', and :84 records demo-hp was reprovisioned (address 192.168.0.87 -> 192.168.0.104, read 2026-09-21). Whether the reprovisioned box is actually on the island is a live-box fact (agent.json, pct config) not provable from source.", "dup_of": null, "unique_facts": "A reprovision since the row was filed (fresh installs are born-on-island per memory) may have made nodes.md true; a read-only check of demo-hp's agent.json listen_addr settles it.", "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-340", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/scripts/felhom-tenantsync.sh: no health op and no 8007 probe (grep '8007\\|health)' empty). Needs an ep0 (protected) script version bump + hub signal (M).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-349", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-agent/internal/hub/report.go:282-292 reports only WrapperSHA256; no agent binary sha field in the report (grep AgentSHA256 finds only the hub's manifest entry, hub/internal/api/handler.go:2726). R-349 commits 40d857b/910fd911 are the manual correction only.", "dup_of": null, "unique_facts": "Fix spans agent (report own sha) + hub (compare to vouched agent_sha256).", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-350", "sev": "P3", "category": "Security & access", "group": "UNCHECKED", "evidence": "Whether the hub operator password was rotated after 2026-08-20 lives only in the hub DB / operator; git log shows no rotation record (only 910fd911 filing it). Not determinable from source.", "dup_of": null, "unique_facts": "The reusable lesson is already in memory (curl-w-redirect-url-leaks-credentials). If the operator confirms no rotation and accepts the risk (value is in a local transcript on DooPlex only), this can close as accepted; rotation itself costs ~5 minutes via /configuration.", "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-362", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/internal/backup/offbox_restore.go:380, offbox.go:1929, shares_restore.go:116: `return fmt.Errorf(\"restore dir: %w\", err)` — no drive-state consultation on the restore path; IsDisconnected (settings.go:2001) is consulted only by backup legs/update guard (backup.go:609, tier2.go:457, …). No R-362 commit.", "dup_of": null, "unique_facts": "The settings IsDisconnected flag may lag a detach by seconds (the observed detach was 4 s into the restore), so the check should also test the drive mountpoint directly, not only the flag.", "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/backup/offbox_restore.go", "controller/internal/backup/offbox_restore_test.go"], "change": "When MkdirAll/write of the restore destination fails with EACCES/ENOENT, check whether the destination's drive is disconnected/decommissioned or no longer a mountpoint, and return a Hungarian error naming the drive instead of the raw permission error.", "test": "Unit test with a temp dir made read-only plus a settings stub marking the drive disconnected -> error names the drive, not 'permission denied'; red on current code.", "minutes": 60}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-363", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-SMALL", "evidence": "felhom-controller/controller/cmd/controller/main.go:1546: `sched.Daily(\"fill-watch\", \"03:30\", func(ctx context.Context) error { return fillWatcher.Check() })`. Watcher emits on escalation only with a persisted band (internal/fillwatch/fillwatch.go:133, :231), so a faster cadence does not spam.", "dup_of": null, "unique_facts": "Recommendation not followed: sharing the reserve's reading is a design change; an hourly sched.Every gives the same outcome cheaply because Check() is escalation-only and persisted.", "small_fix": {"repo": "felhom-controller", "files": ["controller/cmd/controller/main.go (gitignored dir — git add -f)", "a source-pin test"], "change": "Replace sched.Daily(\"fill-watch\", \"03:30\", …) with sched.Every(\"fill-watch\", time.Hour, …) (scheduler.go:104), keep the startup check.", "test": "fillwatch test: two consecutive Check() calls at the same band emit once (proves hourly is safe); a source-pin test that main.go registers fill-watch via Every.", "minutes": 30}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-388", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Product direction, operator's call; recorded as [DESIGN — DIRECTION] in documentation/architecture/08-alarm-ladder.md §8 per the row. Nothing in source to fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-401", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Event-triggered watch row: felhom-controller/controller/internal/backup/offbox_integrity.go:63 `const integrityCheckTimeout = 30 * time.Minute`, :78 `var integritySlowNoticeThreshold = 5 * time.Minute` — unchanged; the trigger (slow WARN on a large store) has not been recorded.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-409", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller/controller/internal/backup/recovery_unit.go:58 `Checksums map[string]string json:\"checksums\" // sha256 of captured compose/ files`; writers at :147, :152, :202 hash only compose/.felhom.yml/app.yaml — no db_dumps or volume_dumps hash.", "dup_of": null, "unique_facts": "Changes the recovery-unit manifest format (capture + verify side); not under an hour.", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-412", "sev": "P3", "category": "Backup & restore", "group": "NOT-WORTH-IT", "evidence": "Leg 1 shipped: felhom-controller fcef8e0 / 62c6a8a (v0.232.0, R-412a) — hollow push now logs at WARN, pinned by TestR412a_EmptyPushDoesNotReadAsAPlainSuccess. Leg 2 (re-read the unit before push vs accept the race) has no code and is a decision.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A narrow race: if a recovery unit is destroyed inside an off-site run after its own dump leg, the push ships the just-rebuilt hollow unit; the next run repairs it and the WARN line now says it carried no data.", "cost": "A re-read/re-validate step in the push path plus a red-proof drill; touches the capture/push boundary the architecture says not to guard (08 §8.2).", "if_never": "Rarely, one off-site snapshot of one app is hollow until the next nightly run; the R-403 mirror guard keeps the good secondary copy; the WARN makes it visible.", "pick": "close-as-accepted"}, "minutes_spent": 4}
|
||||
{"id": "R-433", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Hetzner answered (ticket per the row): file-level snapshot access is a MAIN-account capability. hub/internal/hetznerapi/hetznerapi.go still has no snapshot read method (only size_snapshots usage, :83-87). The owed proof (read a file from a snapshot with the main account; forced-command append-only key) is a live, credential-bound operator act.", "dup_of": null, "unique_facts": "The row's state cell 'BLOCKED-ON-PROVIDER' is stale — the provider has answered; the row should be restated as 'owed: one main-account measurement' (operator + CC).", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-435", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom.eu/hub/internal/monitor/offsite.go:255: `snapshotDropFraction = 0.5 // more than half the history gone in one step` — no per-tag second signal exists.", "dup_of": null, "unique_facts": "Not in the row: STATUS.md no longer contains a 'noticed within a day' claim (grep -i 'within a day|deletion|half' empty), so that half of the row's reason for staying open is gone; what remains is the per-tag detector (new mechanism).", "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-440", "sev": "P3", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog-felhom.eu @917a779: 15 templates still have no `update_ladder:` in .felhom.yml — bentopdf code-server glance gokapi gramps-web homebox homepage jellyfin onlyoffice plant-it plex recipe-importer seerr vaultwarden wanderer (calibre-web got its first step in 53a4a1d). For these, pins float with no recorded digest.", "dup_of": null, "unique_facts": "Closes per app as R-462 (widen the upgrade harness) proves a first ladder step; the pre-v0.269.0 installs heal on their next guarded update.", "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-444", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No fstrim anywhere in felhom-agent or felhom.eu/scripts (grep 'fstrim' over *.go/*.sh empty). Needs a new periodic host job (agent, privileged, sudoers/bundle) and possibly an operator surface.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-446", "sev": "P3", "category": "App updates", "group": "DUPLICATE", "evidence": "felhom-controller/controller/internal/stacks/updateorder.go:96: `if len(s.CatalogDigests) == 0 || s.CatalogTestedAt.IsZero() { return false }` — blind only for apps with no ladder entry, i.e. the same 15 templates R-440 lists (app-catalog has no update_ladder for them). Both rows close by the same act: each app's first proven ladder step (R-462).", "dup_of": "R-440", "unique_facts": "Move into R-440: the customer-visible consequence — the 'Naprakész' badge cannot go 'behind' for those 15 apps (updateorder.go:96); R-740 (same-tag re-test) is CLOSED (catalog 6a3ead9), so floating-tag drift for laddered apps is handled.", "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-450", "sev": "P3", "category": "App updates", "group": "FIXED-BY-LATER-WORK", "evidence": "The row's only remainder was 'the other ten PostgreSQL apps need two-venue proof'. R-463 CLOSED 2026-09-30 by felhom.eu 25cb3eb9 ('The last six PostgreSQL apps decided'): 8 of 11 moved by the box's own conversion, 3 (zipline, adventurelog, immich) stay by decision 42; CLOSED-ITEMS.md:223. Source proof: app-catalog templates/docmost/docker-compose.yml:62 'image: postgres:18-alpine' (also rallly:67, outline:64, paperless-ngx:101). The per-app engine gate stays as the permanent rule (catalog CLAUDE.md:115-122).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-458", "sev": "P3", "category": "App updates", "group": "NOT-WORTH-IT", "evidence": "Still true by design: Syncer.copyTemplates copies .felhom.yml verbatim. The row narrows the risk to type: api probes with expect; those exist (e.g. templates/dawarich/.felhom.yml:104 'expect:', adventurelog:93, immich:116, nextcloud:108). Quick catalog history scan (commits touching .felhom.yml health/expect lines AND an image: line) found only new-app commits (96829d0, 72247a3, 882ac14, 195129c, 4351d08) plus lifecycle/re-pin commits a325416/b3eabfd; b3eabfd's wanderer diff showed no healthcheck lines. So no evidence of a healthcheck changed together with a version move on an existing app.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A frozen (pinned-behind) app can receive a newer health check from .felhom.yml. The only result is a false 'degraded'/dead-app alarm, never data loss, and only for type: api probes with an expect block.", "cost": "Freezing just the healthcheck key means the sync must parse and rebuild a metadata file on the path that touches every app every 15 minutes. That is new surface on a hot path.", "if_never": "Possibly a false alarm one day for an app pinned behind, if a catalog commit changes an expect-probe together with a version. History shows none so far, and the ladder now moves apps step by step.", "pick": "close-as-accepted"}, "minutes_spent": 12}
|
||||
{"id": "R-462", "sev": "P3", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Ongoing multi-session work (fixtures and ladders per app). Last progress: catalog e6f3ec2 (2026-09-30), audits/more-night-apps-2026-09-30/. 21 apps still have no ladder (per the row). Not checkable as done from source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-468", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "A standing pre-customer arrangement, not a defect. The mechanism is live: felhom.eu/scripts/golden_currency_gate.py:137 reads documentation/tests/golden-waiver.yml. The waiver file is currently ABSENT: deleted in felhom.eu 5efe6dae (2026-10-04, 'golden 0.292.0 vouched ... golden waiver deleted'). The row retires only at the first external install, so it stays as a watch.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-469", "sev": "P3", "category": "App updates", "group": "STILL-TRUE-SMALL", "evidence": "What stood between this row and its close was R-463, now CLOSED (felhom.eu 25cb3eb9; CLOSED-ITEMS.md:223). The per-app PostgreSQL gate (decision 35) is now the permanent rule. One thing still contradicts source: app-catalog-felhom.eu/CLAUDE.md:103 heading reads '- **A MariaDB major gets its OWN EDGE; PostgreSQL and MySQL may not cross a major at all.**', but the body at :115-122 lets PostgreSQL cross per app with engine_conversion + two-venue proof, and 8 apps already did (e.g. templates/docmost/docker-compose.yml:62 'image: postgres:18-alpine').", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["CLAUDE.md"], "change": "Reword the rule heading at CLAUDE.md:103 and the 'What is NOT lifted' paragraph (:112-114) to say that PostgreSQL crosses a major one app at a time with engine_conversion + both-venue proof (decision 35) and that MySQL stays refused. Then close R-469 citing R-463's closure. Do not delete the gate: it is now the per-app enforcement.", "test": "Run python3 scripts/catalog_gates.py (unchanged result expected). grep CLAUDE.md for 'may not cross a major at all' -> 0 hits.", "minutes": 20}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-489", "sev": "P3", "category": "Apps & catalog", "group": "FIXED-BY-LATER-WORK", "evidence": "The residual (a unit-restore-recreated volume has no compose label, so the remove answered []) was fixed in felhom-controller 206b035 (v0.268.0, R-658). controller/internal/stacks/delete.go:1089-1090: '// appVolumeSet is every volume the removal accounts for: the ones carrying the project label AND the // ones the app's definition declares that Docker holds by name (R-658, v0.268.0).' delete.go:703 'resp.VolumesRemoved = removedVolumes(volsBefore, m.appVolumeSet(name, stackDir))'.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-498", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "Still literal: grep -l '\\.DOMAIN' over app-catalog templates/*/.felhom.yml -> 58 of 58; e.g. templates/bookstack/.felhom.yml:73 \"- 'Nyisd meg a wiki.DOMAIN címet a böngészőben'\". The controller renders first_steps as-is: controller/internal/web/templates/app_info.html:224 '{{range .AppInfo.FirstSteps}}<li>{{.}}</li>{{end}}' (also deploy.html:684). The precedent substitution exists for default creds only: controller/internal/web/known_login.go:67 'strings.ReplaceAll(meta.AppInfo.DefaultCreds, \"DOMAIN\", s.cfg.Customer.Domain)'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/handlers.go (app info view model)", "controller/internal/web/known_login.go (pattern)", "new test in controller/internal/web/"], "change": "When building the app info view, rewrite each first_steps entry: replace '<word>.DOMAIN' with the stack's real address (installed SUBDOMAIN + customer domain when installed, else the template default subdomain + customer domain). Use the same approach as known_login.go:67.", "test": "A render test over every catalog template's first_steps (or a fixture of 3) asserting that no rendered string contains the literal 'DOMAIN'. A red-proof with the substitution removed must fail.", "minutes": 60}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-516", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-NOT-SMALL", "evidence": "By its own text it now waits for a Hungarian walk on a box with a second drive (items 4, 7, 8, 9, 10) and needs a separate row for item 11. That is live-box work, not source. No commit after 2026-09-20 names R-516 as closed.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-521", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No storage-disconnect suppression of app_start_failed: controller/cmd/controller/main.go:2796 'down := (stacks.IsDownState(st.State) || crashLooping) && !userStopped && !quiesced[st.Name]' (only user-stop/quiesce suppress), and controller/internal/notify/notifier.go:718 emits app_start_failed per newly-down app. It also needs the hub cooldown semantics changed (per-key cooldown outliving storage_reconnected) and an operator decision on customer mail policy. That is two repos plus a decision.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-522", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No tunnel-connection signal in the controller: grep for TunnelConnected/cloudflared connection state in controller/internal -> none. The tile comes from container metadata: controller/internal/web/inframeta.go:21-22 '\"cloudflared\": { DisplayName: \"Cloudflare Tunnel\"'. Fixing it needs a new state source (cloudflared metrics or the hub-push result) wired into the tile, plus a live internet-cut validation.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-531", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Both measurements are done (audits/evidence-drill-0243-2026-09-16/). What remains is an operator design question about the crash-loop budget (a slow loop every 20 min is never paused). That is an operator decision, not code.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-540", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still a single pool box: felhom.eu/hub/cmd/hub/main.go:367 'poolBoxID, _ := strconv.ParseInt(os.Getenv(\"HETZNER_POOL_BOX_ID\"), 10, 64)'. It needs a selection-rule design and eventually a second box (money). No risk today (0.3 % full per the row).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-542", "sev": "P3", "category": "Storage & devices", "group": "UNCHECKED", "evidence": "The controller passes the agent's 'initialize' list through untouched: controller/internal/web/agent_disk_handlers.go:160-162 mergeAttachCandidates only appends to Attach. The agent puts every candidate under initialize: felhom-agent internal/localapi/disks.go:438 'initialize = append(initialize, c) // every unclaimed disk can be initialized'. Whether a REGISTERED in-guest drive still counts as 'unclaimed' depends on the agent's ListCandidateDisks claim state on a real box (internal/storage/candidates.go). Only a live box shows that. No commit names R-542.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-545", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still no un-configure route: controller/internal/web/server.go:744-783 lists /backup/offbox/{config,toggle,enable-all,offer-dismiss,run,reset,status,restore,place,reconstitute,verify-copy/delete,confirm-escrow,inject-password}; nothing removes a target. The fix is a new destructive customer action (shred data/offbox/) with a hub-escrow refusal check (R-241 rule), Hungarian/English copy and a UI. That is more than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-547", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-SMALL", "evidence": "Still daily: controller/cmd/controller/main.go:1546 'sched.Daily(\"fill-watch\", \"03:30\", func(ctx context.Context) error { return fillWatcher.Check() })' plus one startup check (:1557-1566). The check is edge-triggered against PERSISTED state (comment at :1553-1555), so running it more often adds no repeat mails.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/cmd/controller/main.go", "felhom.eu/documentation/architecture/08-alarm-ladder.md (one sentence, optional)"], "change": "Also schedule the fill-watch on an interval (e.g. sched.Every(\"fill-watch-fast\", 15*time.Minute, ...) calling the same fillWatcher.Check) next to the daily run, and log the cadence. Alternative if the operator prefers: state in 08-alarm-ladder.md that a transient full disk is out of scope.", "test": "Unit test with a fake usage func: a target that crosses 95 % between two interval ticks yields exactly one notification, and a second tick at the same level yields none (pins edge-triggering under the faster cadence).", "minutes": 50}, "not_worth": null, "minutes_spent": 8}
|
||||
{"id": "R-548", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "The row's original fix shape SHIPPED: felhom-agent 0722b2c (2026-09-24, R-685) checks before start, internal/backup/runner.go:279 '... so a new one needs about %s (old archives are removed only after a successful backup)'. Shown in the UI by controller 44ae4de (v0.272.0). The 2026-09-30 addendum is still true and is the open part: the local tier is refused for ever (10 refusals on demo-hp) because the old archive is removed only after a success. Nothing gives the room back. That needs a design (prune before write, or move the tier), so it is not small. Consider re-scoping the row to that residual.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-552", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "The interrupted-restore notice is cleared only at controller/internal/backup/opstatus.go:61 'delete(m.opInterrupted, stack)' inside BeginRestoreOp. removeStack (controller/internal/api/router.go:955) clears only the update hold, at router.go:1054 'if cleared, err := r.sett.ClearUpdateHold(name); err != nil {'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/backup/opstatus.go", "controller/internal/api/router.go", "tests in internal/backup and internal/api"], "change": "Add Manager.ClearInterruptedRestore(stack) (delete from opInterrupted + persistRestoreRecordLocked under m.mu). Call it in removeStack next to ClearUpdateHold, and log when it cleared something.", "test": "Unit test: record an interrupted restore for app X, call ClearInterruptedRestore(X), and assert that the list is empty and the persisted record no longer holds X. Wiring test: removeStack on a router with a fake backup manager calls the clear. Red-proof by removing the call.", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-554", "sev": "P3", "category": "Install & onboarding", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still present: controller/internal/setup/ (setup.go, handlers.go, csrf.go, network.go, templates/), and controller/cmd/controller/main.go:328-330 'if setup.NeedsSetup(cfg) { ... runSetupMode(cfg, logger)'. The fix deletes a package and adds a new waiting page (HU+EN copy) with a red-proof. It also needs a check of drill/golden reliance on .needs-setup (controller/internal/web/handler_debug.go references it). More than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-562", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Needs an operator word on the Hungarian number/date format (the row says so) and a deliberate Hungarian-byte change release with parity re-capture. Not checked further.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-565", "sev": "P3", "category": "Process & tooling", "group": "STILL-TRUE-SMALL", "evidence": "The English page test detects only accented letters: controller/internal/web/i18n_parity_test.go:563 'func huLetter(s string) bool {' used by TestI18nEnglishPages (:601). The ASCII Hungarian list exists only in the extractor: controller/scripts/i18n_extract.py:40 'ASCII_HU = re.compile(r\"\\b(Fut|Nincs|Igen|Nem|Hiba|Mentve|...'. No Go test uses it.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/i18n_parity_test.go"], "change": "Add an ASCII Hungarian word regex to TestI18nEnglishPages (seeded from i18n_extract.py ASCII_HU plus the words releases B/C found: mp, db, FIGYELEM, jelenlegi, majd a(z), Konfig, Megtartva, helyi, Befejezve, automatikus, kedd/szerda/szombat, szint), applied after the data mask.", "test": "Negative control: a plain English sentence passes. Decoy: plant 'mp' in an English value and the detector must flag it. Then run the existing English renders (go test ./internal/web -run TestI18nEnglishPages, no Docker).", "minutes": 55}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-573", "sev": "P3", "category": "Monitoring & notifications", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom-controller 7c4a33b (v0.258.0, 'the last four Hungarian things an English household met ... R-573 the two channel banners'). controller/internal/web/alerts.go:113 'func (am *AlertManager) SetAgentChannelAlert(down bool, msgKey, msg string) {' and :136 'func (am *AlertManager) SetEndpointDriftAlert(drift bool, msgKey, msg string) {'. Both set MessageKey with msg only as a fail-open fallback.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-575", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "Still a plain string: controller/internal/stacks/deploy.go:1456 'func (m *Manager) memoryVerdict(newReqMB, newLimitMB, releasedReqMB, releasedLimitMB int) (refusal error, warning string) {'. Callers are deploy.go:302 and update.go:497. The consequence is stated at controller/internal/stacks/deploy_errors.go:41.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/stacks/deploy.go", "controller/internal/stacks/update.go", "the deploy response renderer in internal/web or internal/api"], "change": "Make memoryVerdict return (refusal error, warningKey string, warningArgs []any) instead of a msgHU string, and render the warning at the deploy answer with the request's language (the Alert/UpdateRefusal pattern).", "test": "Test that a deploy over the soft memory line returns the English warning for lang=en and the unchanged Hungarian bytes for hu (parity fixture). Red-proof: forcing hu rendering fails the en case.", "minutes": 60}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-578", "sev": "P3", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "The lock-reentrancy guard is still one test in one package: controller/internal/backup/offsite_diag_test.go:190 'TestNoteHelpersAreNotCalledUnderTheSettingsLock'. No other package has one, and there is no gate (grep for R-578 in controller -> none). The fix needs a cross-package AST gate that knows which methods read settings (or a re-entrant read path in Settings). That is a new mechanism, likely more than an hour with decoys.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-581", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "Still no tie-break: felhom.eu/hub/internal/store/store.go:1329 'SELECT customer_id, MAX(received_at) as max_time' joined on 'r.received_at = latest.max_time' (:1333). A same-second tie returns BOTH rows (a duplicate customer in GetCustomers), not just an arbitrary one. Other newest-report queries also order on received_at alone: store.go:1393 and :1446 'ORDER BY received_at DESC'. Compare :3574, which already has ', id DESC'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/internal/store/store.go", "hub/internal/store/*_test.go"], "change": "In GetCustomers, join on MAX(id) per customer_id instead of MAX(received_at), and add ', id DESC' to the ORDER BY at store.go:1393 and :1446.", "test": "Store test: write two reports for one customer in the same tick with different health/version, then assert that GetCustomers returns exactly one row for that customer and that it carries the second report. Red-proof against the current query (expect 2 rows or the wrong one).", "minutes": 40}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-584", "sev": "P3", "category": "Security & access", "group": "NOT-WORTH-IT", "evidence": "Still true as a process gap: the only written rule is felhom-controller/.claude/rules/ui-hungarian.md (credential-bearing helper cleanup). grep for 'shred' over the workspace .claude/rules and felhom.eu/skills -> no hits, so there is no 'last act of a phase' mechanism. The five files were already shredded (per the row).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Helper scripts with the shared DEMO controller password inline were left in a demo guest's /tmp. The cleanup rule exists, but no mechanism enforces it.", "cost": "A real mechanism would be a push-helper wrapper that writes credentials to a 0600 file and shreds on exit, and every session would have to use it. Another rule line would be a wish, as the row itself says.", "if_never": "Demo-box (Tier 0) password litter may recur on throwaway guests. The password is a shared demo one, not a customer one, and rotating it is cheap.", "pick": "close-as-accepted"}, "minutes_spent": 5}
|
||||
{"id": "R-585", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still finished Hungarian from callers: controller/internal/notify/notifier.go:408 'n.PushEvent(\"backup_failed\", \"error\", message, BackupDetails{Error: errMsg})', :487 offbox_enlarge_blocked, :497 db_dump_failed. None of the six types is in convertedProducers (controller/internal/notify/message_customer_test.go:39). The hub still has no customerMessages entry for offbox_enlarge_blocked (felhom.eu/hub/internal/notify/templates.go:94/153). The fix changes six producers' callers across packages: more than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-586", "sev": "P3", "category": "Process & tooling", "group": "NOT-WORTH-IT", "evidence": "The harness is still in no push-time gate or CI: grep for 'bootstrap-modes' over felhom.eu/scripts and .gitea -> none. It is required only in documentation/runbooks/iso-release-gate.md:285 'docker run --rm -v <repo>/scripts/iso:/work felhom-iso-assistant:trixie bash /work/test/bootstrap-modes.sh'. The FIFO fix shipped per the row.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The ISO bootstrap harness runs only at each ISO release gate (G16), not on every push.", "cost": "CI needs a container-capable job (a new runner capability), or the push hook needs Docker. Both are new infrastructure for a harness that matters only when an ISO is released.", "if_never": "A harness break is caught at the next ISO release instead of at the push. ISO releases are infrequent, and that gate already blocks a broken one.", "pick": "close-as-accepted"}, "minutes_spent": 6}
|
||||
{"id": "R-587", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "The files are gone (ls /mnt/5_hdd/felhom.eu/felhom-iso/out | grep -c rootpw -> 0). The guard is still not built: the publish still relies on the include pattern, felhom.eu/skills/felhom-build-deploy/SKILL.md:115 'rclone/rclone:latest copy /data R2:felhom-iso --include \"felhom-installer-<VER>*\"'. The build still emits the file for appliance mode: felhom.eu/scripts/iso/build-felhom-iso.sh:351 '... > \"$OUT_ISO.rootpw.txt\" )'.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu", "files": ["scripts/iso/build-felhom-iso.sh", "skills/felhom-build-deploy/SKILL.md (publish step)", "scripts/iso/test/ (a small test)"], "change": "At the start of build-felhom-iso.sh, refuse (non-zero exit, named reason) when any *.rootpw.txt exists in the output dir. In the SKILL publish block, add a pre-check line that aborts the rclone copy if a *.rootpw.txt is present in the publish source.", "test": "Shell test: a temp out dir with a decoy x.rootpw.txt makes the build entry refuse with exit != 0. The same dir without it proceeds past the check (stub the rest with an early-exit env flag).", "minutes": 45}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-593", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-SMALL", "evidence": "Still wrong: app-catalog-felhom.eu/templates/papra/.felhom.yml:47 ' description: \"Az alkalmazás aldomainje\"' sits under AUTH_SECRET (:37). SUBDOMAIN (:30-35) has no description.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "app-catalog-felhom.eu", "files": ["templates/papra/.felhom.yml", "scripts/copy_freeze/hu.json"], "change": "Move 'Az alkalmazás aldomainje' to SUBDOMAIN, give AUTH_SECRET its own description (e.g. 'A munkamenetek aláírásához használt kulcs — ne generáld újra'), add the two English i18n.en descriptions, and re-capture papra's entries in copy_freeze/hu.json with the reason in the commit.", "test": "python3 scripts/catalog_gates.py green (copy-freeze and i18n coverage). Catalog English ceiling for papra reaches 14/14.", "minutes": 25}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-600", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "Log line unchanged: felhom.eu/hub/internal/web/customer_delete.go:310 's.logger.Printf(\"[INFO] customer DELETE cascade COMPLETE for %s (journal #%d) — full teardown\", customerID, journalID)'. The wgsync Trigger is called only from hub/internal/api/wg.go:86 'h.wgSyncer.Trigger()', not from the customer cascade. Measured lag is seconds to about 6 min (row).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": ["hub/internal/web/customer_delete.go", "hub/cmd/hub/main.go (wire the reconciler into the web server if not already)", "test in hub/internal/web"], "change": "Call the wgsync reconciler's Trigger() before the COMPLETE log line (nil-safe), and make the line say 'wg peer removal pushed' or 'queued for the next wgsync push' depending on whether a syncer is wired.", "test": "Web test with a fake reconciler: a completed delete cascade calls Trigger exactly once, and the log line contains the peer-removal state. Red-proof by removing the call.", "minutes": 50}, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-607", "sev": "P3", "category": "App updates", "group": "UNCHECKED", "evidence": "The row asks first for a live reproduction loop (push a tag, sync, read catalog_images on a timer) and has no diagnosis. Why the sync reports 'nincs változás' while the cache moved, and when CatalogImages refreshes, are live-box behaviour I could not settle from source in the time box. No commit after d19f07ea (filing) names R-607 as fixed.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-612", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "The memory half is fixed (catalog a5a729a, 'wishlist 512M (R-612)'). The open half, making a failed first-boot seed visible, needs a new detection mechanism (read the seed's exit/log, or a probe that checks the Role/Group rows). No commit addresses it.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-613", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "uptime-kuma is fixed (catalog a5a729a). The open half is a sweep of all 58 templates for probes that pass on a setup wizard. That needs per-app live inspection, so it is not small. No commit names it.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-615", "sev": "P3", "category": "App updates", "group": "STILL-TRUE-SMALL", "evidence": "Still inert: controller/internal/sync/sync.go:279 clones only 'if _, err := os.Stat(gitDir); os.IsNotExist(err)'. Later cycles run :299 'fetch --depth 1 origin' against the stored remote, and nothing compares cfg.Git.RepoURL with origin (grep 'set-url|remote' in sync.go -> none).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/sync/sync.go", "controller/internal/sync/sync_test.go"], "change": "Before the fetch, run 'git remote set-url origin <buildRepoURL()>' (or read origin and re-clone on mismatch), and log at INFO, masked, when the remote changed. The appended drill-folder residue (stack folders the live catalog lacks) is a separate drill-teardown item and is not part of this fix.", "test": "Test with two local bare repos (git only, no Docker): clone A, switch cfg.Git.RepoURL to B, sync, and assert that the cache HEAD is B's commit. Red-proof against current code.", "minutes": 50}, "not_worth": null, "minutes_spent": 7}
|
||||
{"id": "R-616", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-SMALL", "evidence": "Still true: controller/internal/sync/sync.go:327-331 buildRepoURL injects 'https://%s:%s@' and the clone at :283-288 passes that URL to 'git clone', so it persists as origin. maskRepoURL (:100) masks only log lines. Inert on the fleet while git.token is empty (row).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/sync/sync.go", "controller/internal/sync/sync_test.go"], "change": "Clone and fetch with the credential-free RepoURL, and supply credentials per command via '-c http.extraHeader=Authorization: Basic <b64>' (or GIT_ASKPASS env) only when username+token are set. Pairs naturally with the R-615 set-url fix (set-url to the bare URL).", "test": "Test with a local repo and a token configured: after clone and after a sync, 'git -C cache config remote.origin.url' contains no '@'. Red-proof on current code. The operator token rotation stays a separate operator act.", "minutes": 55}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-622", "sev": "P3", "category": "App updates", "group": "FIXED-BY-LATER-WORK", "evidence": "adventurelog v0.13.0 was diagnosed, fixed and promoted with a two-venue test record in app-catalog-felhom.eu 06ea7da (2026-09-27, 'adventurelog: v0.12.1 -> v0.13.0 with its health and world-data fixes in the same commit (R-655, 09 decision 41)'; bench healthy in 217 s, box 9202 through the guarded Update in 204 s). templates/adventurelog/docker-compose.yml:13 ' image: ghcr.io/seanmorley15/adventurelog-backend:v0.13.0'. The proven step is recorded at templates/adventurelog/.felhom.yml:132 (update_ladder entry from v0.12.1).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-635", "sev": "P3", "category": "Monitoring & notifications", "group": "FIXED-BY-LATER-WORK", "evidence": "The open remainder (app_oom fires once per container run, no escalation) was built in felhom-controller 0054d4b (v0.265.0, 'OOM storm alarm', R-636). controller/internal/notify/notifier.go:746 '\\t\\tn.emit(\"app_oom_storm\", \"error\",' fires once per run when 20 or more kills land in 30 min (:754-764, oomStormKills=20, oomStormWindowMin=30; pinned by TestR636_*). The 79 % headroom and the method lesson are carried by R-462 (per the row).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-645", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still true for the operator CLI: controller/internal/settings/settings.go:1923-1929 ClearRestoreHold deletes ANY hold reason (including update-failed) with no pin restore, and the flag is in controller/cmd/controller/main.go:94/198. The row lists three candidate shapes with 'none chosen'. Picking one changes operator-path semantics and the capture logic, so it is not a one-hour fix. (The automatic undo is no longer affected, since v0.263.0, per the row.)", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 6}
|
||||
{"id": "R-675", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-SMALL", "evidence": "Unchanged: controller/internal/web/handlers.go:1745 'return head + \"A fájlok a második meghajtó másolatából állíthatók vissza: „Fájlok visszaállítása”.\"'. The branch does not check for a whole copy on the second drive (decision 26, v0.269.0).", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom-controller", "files": ["controller/internal/web/handlers.go", "a handler test in controller/internal/web"], "change": "In missingFileLegsRefusal, when the second drive holds a whole copy of the app (the decision-26 whole-copy check the backup manager already exposes for the restore page), name that whole restore action instead of 'Fájlok visszaállítása'. Keep the existing branch otherwise.", "test": "Three-branch test with a fake backup manager (off-site row / whole copy on drive 2 / tier-2 files only / nothing), asserting each sentence. Red-proof: the whole-copy case fails on current code.", "minutes": 45}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-676", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "A watch row, still true: controller/internal/stacks/unhealthy.go:116 skips only 'if st.Deploying || st.Updating || st.HoldReason != \"\" || st.updateHeld {', so a deploy's first start (after Deploying clears) is sampled by decision 28's crash-loop stop. The immich cause is fixed in the catalog (56c4888, 768M, per the row). Covering a slow first start would need a first-start grace decision.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-682", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "felhom-controller 7690c27 (v0.296.0): no remove journal in source — grep -rni 'remove.*journal|removeJournal|remove_intent|interrupted remove' controller/*.go returns nothing; git log --grep R-682 empty. Needs a new boot-time journal mechanism.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-683", "sev": "P3", "category": "App updates", "group": "UNCHECKED", "evidence": "Watch item about a power-cut drill outcome; behaviour only a live box shows; git log --grep R-683 empty in controller.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-698", "sev": "P3", "category": "Backup & restore", "group": "NOT-WORTH-IT", "evidence": "Row is an open operator decision among options (a)-(d); no source change implied. No commit references R-698 fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "A backup records the image name/digest, not the image; restoring a version the maker deleted from the registry fails at the pull.", "cost": "Options b-d add a new mirror or tens-hundreds of MB per app per copy on every tier, a new part on the recovery path.", "if_never": "A restore of a deleted version fails; the household uses the next copy or a newer version (option a). All 42 ladder digests resolved when measured.", "pick": "close-as-accepted (option a), operator to confirm"}, "minutes_spent": 2}
|
||||
{"id": "R-700", "sev": "P3", "category": "Storage & devices", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom-controller 820e8ef (v0.276.0, R-697/R-700). controller/internal/stacks/migrate.go:789: 'm.logger.Printf(\"[INFO] [stacks] %s: data moved %s -> %s — app.yaml keeps its pin (%d service(s)) and records\"' — persistDriveFlip (migrate.go:763) loads app.yaml and changes only HDD_PATH. Only a live proof on a two-drive box remains (row's own residue).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-704", "sev": "P3", "category": "Apps & catalog", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom-controller 7cba0bf (v0.278.0). controller/internal/api/router.go:918 'func (r *Router) dropLeftoverHold(name, why string) {' calling r.sett.ClearUpdateHold(name); pinned by TestR704_AFreshInstallDropsALeftoverHold. Residue: live proof of the install-time drop only.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-706", "sev": "P3", "category": "Backup & restore", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom-controller 0c702f8 (v0.279.0). controller/internal/api/router.go:946 'if err := r.backupMgr.DeleteOffsiteRestoreCopy(name); err != nil {' inside removeVerificationCopy (R-706); pinned by TestR706_RemovalWithBackupsDeletesTheVerificationCopy. Residue: not seen live.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-717", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "app-catalog templates/opengist/.felhom.yml:42 and templates/wishlist/.felhom.yml:42 carry only signup_block; no after_setup/after_install in either .felhom.yml (grep empty). Fix needs per-app DB writes (opengist sqlite with app stopped) plus live proof.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-723", "sev": "P3", "category": "Monitoring & notifications", "group": "FIXED-BY-LATER-WORK", "evidence": "felhom.eu 80aeac71 (hub v0.126.0). hub/internal/monitor/staleness.go:174 '// R-723 (v0.126.0): a customer's NEW box is not a recovery.'; hub/internal/notify/dispatcher.go:540 '\"suppressed\", \"first hour of a new box (R-723)\", \"operator\"'. Residue: live proof at a real first install.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-724", "sev": "P3", "category": "Monitoring & notifications", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Text parts fixed in controller 6be6c53 (v0.283.0). Remaining LAN/gateway read still goes only through the samba container: controller/internal/stacks/guestnet.go:47 'out, err := dockerexec.Command(\"docker\", append([]string{\"exec\", sambaContainer}, args...)...).Output()' — the comment (guestnet.go:18) accepts that reads fail while sharing is off. Needs another read path (design).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-728", "sev": "P3", "category": "Hub & operator", "group": "STILL-TRUE-SMALL", "evidence": "No commit fixes R-728 (git log --grep in felhom.eu only the filing commit 11591f3a). hub/internal/web/configs.go:705 'existing, _ := s.store.GetCustomerConfig(customerID)' is a check-then-act before slow applyOffsite/applyPBSDR; store.go:1592 save is an upsert ('retrieval_password = excluded.retrieval_password'), so two concurrent submits both pass and both mint.", "dup_of": null, "unique_facts": null, "small_fix": {"repo": "felhom.eu (hub)", "files": "hub/internal/web/configs.go (+ a new _test.go)", "change": "Add a per-customerID in-flight guard (sync.Map/mutex set) around the create handler from the duplicate check to the self-bind mint, so a second concurrent submit for the same ID gets the 'already exists' form; optionally disable the submit button on submit in the template.", "test": "Unit test firing two concurrent POSTs for the same customer ID with a blocking applyOffsite seam; assert exactly one 'Customer config created' / one self-bind mint; red-proof by removing the guard.", "minutes": 60}, "not_worth": null, "minutes_spent": 5}
|
||||
{"id": "R-729", "sev": "P3", "category": "Backup & restore", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No route clears the off-site target: controller/internal/web/server.go:744-783 lists /backup/offbox/{config,toggle,enable-all,offer-dismiss,run,reset,status,restore,place,reconstitute,verify-copy/delete,confirm-escrow,inject-password}; /reset (offbox_handlers.go:311) only resets an orphaned repo. New press needs handler + settings clear + template + HU/EN copy + escrow/hub-managed-target interplay — more than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-733", "sev": "P3", "category": "Process & tooling", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Harness/golden-evidence change plus a decision whether proofs run with swap off; no commit references R-733. Not a source-verifiable single fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-738", "sev": "P3", "category": "App updates", "group": "NOT-WORTH-IT", "evidence": "wger part FIXED: app-catalog 7a4ff48 'wger: run the database migrations at start (R-738)'; templates/wger/docker-compose.yml:46 ' - DJANGO_PERFORM_MIGRATIONS=True'. Only the residue stays: the guarded Update's health check reads only the front page.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The guarded Update's health check sees only an app's front page, so an app that serves its front page while its data is broken passes as done.", "cost": "A per-app data read-back inside the product's update check — a new mechanism across every template (the harness fixture already does this off-box).", "if_never": "Catalog onboarding's fixture read-back keeps catching such breaks before a version reaches the ladder; a box could still report done on a broken update the catalog did not test.", "pick": "close-as-accepted (wger defect fixed; the generic gap is a design note)"}, "minutes_spent": 3}
|
||||
{"id": "R-747", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Lockout shortened: app-catalog a4597cd; templates/mealie/docker-compose.yml:27 ' - SECURITY_USER_LOCKOUT_TIME=1'. Residue still open: hourly lock renewal by a stranger (needs decision 57 option d) and page copy; needs decision + live proof.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-755", "sev": "P3", "category": "Apps & catalog", "group": "DUPLICATE", "evidence": "Still true: templates/wger/docker-compose.yml has no WGER_USE_GUNICORN (grep empty). R-762 (open, read) states 'Owner decides together with R-755 (same server question)' and its fix names 'the gunicorn switch of R-755'.", "dup_of": "R-762", "unique_facts": "wger runs `manage.py runserver` (Django dev server) measured via ps on 9202 2026-10-01; upstream entrypoint.sh:81-87 starts gunicorn only with WGER_USE_GUNICORN=True; the fix needs its own memory watch on bench + box.", "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-756", "sev": "P3", "category": "Storage & devices", "group": "UNCHECKED", "evidence": "Depends on whether 9202's scratch drive is a registered drive — live box state; the row itself says not measured which. Not verifiable from source.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-757", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "No commit references R-757. controller/internal/stacks/deploy.go:1339-1349: 'case \"secret\":' ... 'value, err := generateValue(field.Generate)' ... 'appCfg.Env[field.EnvVar] = value' for any missing field of a deployed app, with no exception for fields consumed only by after_install. Fix needs a design (a marker for given-at-install fields or an ask path).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-758", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Still true: templates/bookstack/.felhom.yml:18 ' mem_limit: \"512M\"' vs compose limits docker-compose.yml:42 '512M' + :79 '256M'; onboarding/EXISTING-APPS-GAPS.md:26 still lists all 8. No gate (only scripts/onboarding_gaps.py:171 reports it). Not small: raising 8 figures changes the capacity check (decision 22) — which apps fit a box — plus a gate with decoy and a publish.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-762", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "templates/wger/docker-compose.yml sets no DJANGO_DEBUG and no static/media server (grep empty); templates/wger/.felhom.yml:18 'lifecycle: hidden' (catalog 55b8c8a). Needs a server design decision (nginx sidecar vs gunicorn+static) and bench+box proof.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-763", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "templates/wger/docker-compose.yml sets neither ALLOW_REGISTRATION nor ALLOW_GUEST_USERS (grep empty); wger hidden (.felhom.yml:18 'lifecycle: hidden'). The env change is tiny but its proof needs a working wger on 9202 (blocked by R-762); best done in the same session as R-762.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-774", "sev": "P3", "category": "Apps & catalog", "group": "STILL-TRUE-NOT-SMALL", "evidence": "templates/karakeep has no Sentry/phone-app sentence (grep -i sentry empty); mail-ON proof needs a hub-enabled live box (demo-hp) — live work.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-775", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Narrowed (Grimmory published behind the family gate). Residue: per-name 15-min lock is hard-coded upstream (no setting) and the reinstall-over-kept-books finding is uninvestigated — needs live investigation.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 2}
|
||||
{"id": "R-776", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Only bookstack has it: templates/bookstack/docker-compose.yml:32 ' - APP_PROXIES=172.16.0.0/12'; grep for TRUSTED_PROXIES/CORE_TRUST_PROXY/IPEXTRACTION/N8N_PROXY_HOPS in kimai, zipline, vikunja, nextcloud, n8n compose returns nothing. Five apps, each needing a live 3.6 re-measure on 9202.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-778", "sev": "P3", "category": "Security & access", "group": "NOT-WORTH-IT", "evidence": "Reasoned-only row; controller client-address code now in controller/internal/web/clientaddr.go (v0.286+). The window exists only during a self-update crash roll-back below 0.286.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "If a box rolls back to a controller older than 0.286, the dashboard's login counter trusts the leftmost forwarded address and can be dodged until the box moves forward.", "cost": "A 0.285.x patch release or a self-update rule refusing to roll back across 0.286 — release work for a rare window.", "if_never": "The window lasts only from a crash roll-back to the next floor delivery; the floor never moves back. A stranger could make more password guesses during that window.", "pick": "close-as-accepted"}, "minutes_spent": 2}
|
||||
{"id": "R-782", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Source agrees: templates/glance/docker-compose.yml:27 seeds glance.yml with no auth: block (grep 'auth' in templates/glance empty); templates/homepage has no HOMEPAGE_ALLOWED_HOSTS (grep empty). Needs live measurement on 9202 and a decision whether a public glance dashboard is intended.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-783", "sev": "P3", "category": "Security & access", "group": "NOT-WORTH-IT", "evidence": "Measured upstream behaviour (better-auth 3 per 10 s keyed on one address); SparkyFitness exposes no env for ipAddressHeaders. No source change possible in the catalog.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Three wrong SparkyFitness sign-ins by anyone block every visitor's sign-in for about 10 seconds.", "cost": "Needs an upstream setting (better-auth ipAddressHeaders) plus a right-walking reader — not in our control.", "if_never": "A stranger retrying every 10 s can keep the household out of sign-in; one who stops frees it within 10 s. Not forgeable.", "pick": "close-as-accepted (re-open if upstream exposes the setting)"}, "minutes_spent": 2}
|
||||
{"id": "R-785", "sev": "P3", "category": "App updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "templates/sparkyfitness/docker-compose.yml:45 ' image: codewithcj/sparkyfitness_server:v0.17.3' and :89 'codewithcj/sparkyfitness:v0.17.3'. Major-version ladder walk (bench + box), gated on R-784.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-831", "sev": "P3", "category": "Security & access", "group": "NOT-WORTH-IT", "evidence": "Operator decision 73: not rotated by choice. Nothing in source to change.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "The Hetzner storage API token was printed into one session transcript.", "cost": "Three manual console/kubectl steps by the operator (about 10 minutes).", "if_never": "Anyone who obtains that transcript could create, reset or delete Storage Box sub-accounts.", "pick": "keep (rotation is the operator's call; do not close a leaked-secret row silently)"}, "minutes_spent": 1}
|
||||
{"id": "R-836", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Live host boot-loader work needing operator-approved reboots and measurement (GRUB env block on ESP, sp5100_tco arming).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-839", "sev": "P3", "category": "Box system & updates", "group": "STILL-TRUE-NOT-SMALL", "evidence": "Gate behaves as described: controller/cmd/controller/main.go:2397 'return false, \"drive \" + hdd + \" is not a live mountpoint\"' with hdd = cfg.Env[\"HDD_PATH\"] (main.go:2379). Which writer put a per-app path in HDD_PATH is undiagnosed — a diagnosis task, not a one-hour fix.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 4}
|
||||
{"id": "R-853", "sev": "P3", "category": "Box system & updates", "group": "NOT-WORTH-IT", "evidence": "Still true: felhom-agent e06ed97, cmd/felhom-agent/main.go:3765 'if !f.at.IsZero() && time.Since(f.at) < 10*time.Minute {' caches a failed guest read; main.go:853 factsReporter needs firstGuest(px). The delay equals the 15-min host report interval, so dropping the failure cache alone does not shorten it; the real fix is a guest-less host facts mode in the wrapper.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "After a boot the box's versions and crash facts reach the hub up to about 15 minutes late.", "cost": "A new guest-less facts mode in the root wrapper plus agent change and release.", "if_never": "Facts arrive one report later after each boot; nothing is lost (the crash guard keeps 7 days).", "pick": "close-as-accepted"}, "minutes_spent": 4}
|
||||
{"id": "R-862", "sev": "P3", "category": "Box system & updates", "group": "UNCHECKED", "evidence": "Waiting on the operator's by-hand bootstrap on Tester 2 through his tunnel; whether done is live-box state. No commit records it (felhom.eu log since 2026-10-04).", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-870", "sev": "P3", "category": "Security & access", "group": "NOT-WORTH-IT", "evidence": "Operator ruling 2026-10-05 option B: not rotated now. Nothing in source to change.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": {"what": "Tester 1's two Cloudflare tokens (disposable test customer) were printed into one session transcript.", "cost": "About 15 minutes of Cloudflare dashboard + hub edit by the operator.", "if_never": "Someone with the transcript could change DNS or the tunnel of the disposable enkicsifelhom.hu test zone.", "pick": "keep (operator's call; close when Tester 1 is retired)"}, "minutes_spent": 1}
|
||||
{"id": "R-879", "sev": "P3", "category": "Security & access", "group": "STILL-TRUE-NOT-SMALL", "evidence": "hub/internal/store/store.go:166 'retrieval_password TEXT NOT NULL,' and store.go:1586/1592 write retrieval_password and api_key as given; no seal on them (grep seal near these fields empty). Sealing/hashing three tables with migration is a security change, more than an hour.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 3}
|
||||
{"id": "R-882", "sev": "P3", "category": "Hub & operator", "group": "UNCHECKED", "evidence": "Longhorn instance-manager state on DooPlex (Tier 2, forbidden to touch); live-only, owner operator.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-883", "sev": "P3", "category": "Hub & operator", "group": "UNCHECKED", "evidence": "homelab-manifests repo is not in this workspace (ls /mnt/5_hdd/felhom.eu/git shows only app-catalog-felhom.eu, drills, felhom-agent, felhom-controller, felhom.eu); live DooPlex check (kubectl) is out of scope.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
{"id": "R-886", "sev": "P3", "category": "Monitoring & notifications", "group": "UNCHECKED", "evidence": "DooPlex Alertmanager volume ownership; homelab-manifests not in this workspace and live check not permitted.", "dup_of": null, "unique_facts": null, "small_fix": null, "not_worth": null, "minutes_spent": 1}
|
||||
@@ -0,0 +1,325 @@
|
||||
# Burn-down 2026-10-05 — Part A: every P4 and P3 row checked against live source
|
||||
|
||||
Method: 8 read-only checker agents, oldest id first (P4 then P3), each row against `main` source (no machine reached); every FIXED / DUPLICATE verdict re-checked by the session before closing (a sample of 22 cited lines re-grepped; one (R-274) held back as half-checked; four (R-700, R-704, R-706, R-723) moved to the operator list — their code fix is in, only the live observation the row waits for is missing). Raw per-row JSON: `partA-results.jsonl`.
|
||||
|
||||
Groups: DUPLICATE 2, FIXED-BY-LATER-WORK 29, NOT-WORTH-IT 43, STILL-TRUE-NOT-SMALL 129, STILL-TRUE-SMALL 91, UNCHECKED 23
|
||||
|
||||
| Row | Sev | Group | Evidence (abridged) | Min |
|
||||
|---|---|---|---|---|
|
||||
| R-10 | P4 | STILL-TRUE-SMALL | FIX: After the os.Rename in DumpOne, open filepath.Dir(finalPath) and call a best-effort dir.Sync() (log at DEBUG on error), mirroring atomicPromoteTar in backup.go:948. — felhom-controller@7690c27 controller/internal/appbackup/dbdump.go:364 `if err := tmpFile.Sync(); err != nil {` then :390 `if err := os.Rename(tmpPath, finalPath); err != nil {` with no directory Sync after; the twin at controlle | 4 |
|
||||
| R-25 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller@7690c27 controller/internal/web/storage_handlers.go:153 `uuid := resolveEnrollUUID(ctx, agent, device)` still resolves by device PATH after format, then AssignDisk(uuid) at the next step; FormatResult (controller/internal/agentapi/client.go:384-395) carries DurableID only for the confirmation path, not the new fs UUID. Binding resolve+assign to the format's durable-id needs the a | 6 |
|
||||
| R-76 | P4 | UNCHECKED | Behaviour is FileBrowser-image runtime behaviour (mode/setgid of UI-created folders), only observable on a live box. The image has changed since the finding: controller/internal/infra/infra.go:27 `FileBrowserImage = "gtstef/filebrowser:1.5.6-stable"` (finding was on 1.3.3). The comment at infra.go:207-208 still asserts `umask 002 so folders the customer creates here come out group-writable (2775 w | 5 |
|
||||
| R-89 | P4 | STILL-TRUE-NOT-SMALL | No retention policy object in hub: `grep -rln -i 'retentionpolicy/retention_policy' felhom.eu/hub` returns nothing (felhom.eu@53d8131b). Commercial per-customer policy = money/product decision + new reconciler. | 2 |
|
||||
| R-91 | P4 | UNCHECKED | Whether /srv/pbs-felhom still exists on ep0 is live-only (ep0 is protected; not touched). Source-side: CONTEXT.md:3656 still reads "`/srv/pbs-felhom` is 13 G of dead weight on `/` awaiting R-91's go-ahead". Extra fact found: documentation/runbooks/offsite-endpoint.md:24 still says the datastore `felhom-offsite` is at `/srv/pbs-felhom` and :119 `proxmox-backup-manager datastore create felhom-offsit | 5 |
|
||||
| R-92 | P4 | STILL-TRUE-SMALL | FIX: Add an exact-bytes value to the PBS DR view (e.g. UsedBytesExact rendered as a title= tooltip or a MB-precision string below 10 GB) without changing fmtBytesGB for other callers. — felhom.eu@53d8131b hub/internal/web/pbsdr_box.go:57 and :64 `view.UsedStr = fmtBytesGB(snap.UsedBytes)`; hub/internal/web/offsite_box.go:54 `return fmt.Sprintf("%.1f GB", float64(b)/float64(int64(1)<<30))` — still | 4 |
|
||||
| R-93 | P4 | NOT-WORTH-IT | PICK close-as-accepted (operator word needed: close, or reopen as 'build a drift fixture'); also drop the dead drill-r50 fence in target-selection.md:111 at close: A row about choosing between two fixtures, neither of which exists any more. | 4 |
|
||||
| R-99 | P4 | STILL-TRUE-NOT-SMALL | No phantom-snapshot cleanup in felhom.eu/hub or felhom-agent (grep -i phantom finds only agent runner/test detection code; no removal path). Deletion on a customer datastore is a separate operator ruling per the row — customer data. | 3 |
|
||||
| R-104 | P4 | STILL-TRUE-SMALL | FIX: Add an OffsiteFailLocked class matched by offboxLockRe in ClassifyOffsiteFailure (before transport) and a cause line in OffsiteFailureMessage telling the operator the repository is locked by an interrupted run and how it clears. — felhom-controller@7690c27 controller/internal/backup/offbox.go:193-222 ClassifyOffsiteFailure has cases NoUnits/NoRepo/Transport and `default: return OffsiteFailUnk | 5 |
|
||||
| R-124 | P4 | NOT-WORTH-IT | PICK close-as-accepted: The disaster-recovery recipe writes the PBS root namespace as the word 'root', but PBS itself uses an empty name, so a pasted '--ns root' fails. | 4 |
|
||||
| R-129 | P4 | STILL-TRUE-SMALL | FIX: After one read-only `ssh -o BatchMode=yes demo-hp true` (and reading root's authorized_keys comment to name the key), rewrite nodes.md 'Access' section to the measured truth and drop the R-129 caveat in target-selection.md:111-112 (also update the memory index line). — Docs still say no key: felhom.eu@53d8131b documentation/operations/nodes.md:110 `### Access — there is no baked SSH key` and | 4 |
|
||||
| R-134 | P4 | STILL-TRUE-SMALL | FIX: Extract a pure zoneCandidates(domain) []string that yields the name and every parent down to two labels, and loop resolveZone over it (same order: most specific first). — felhom.eu@53d8131b hub/internal/cloudflare/unblock.go:117 `for _, name := range []string{domain, parentDomain(domain)} {` and :136-141 parentDomain strips exactly one label (`strings.SplitN(domain, ".", 2)`); controller stri | 4 |
|
||||
| R-161 | P4 | NOT-WORTH-IT | PICK close-as-accepted (residual is a deliberate ruling; owner operator): The runtime check that app data lands on a volume is run by hand, not on every push. | 3 |
|
||||
| R-162 | P4 | NOT-WORTH-IT | PICK close-as-accepted: If Docker ever ran on a storage driver where `docker diff` does not work, the persistence gate would refuse to report and blame the prober instead of the driver. | 3 |
|
||||
| R-164 | P4 | STILL-TRUE-NOT-SMALL | Predicate still absent: felhom-controller controller/internal/appbackup/dbdump.go:544 still only WARNs `its accounts table has NO rows`; restore still replays dump + tar (internal/backup/restore_unit.go:114-118 hasReplayableDump). Blocked on a design (live-vs-dump per-table counts). | 3 |
|
||||
| R-169 | P4 | NOT-WORTH-IT | PICK close-as-accepted (row itself says decide only if the window ever costs something): CI only reports after a push lands, because every repo pushes straight to main with no pull request. | 2 |
|
||||
| R-177 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller@7690c27 controller/cmd/controller/main.go:1546 `sched.Daily("fill-watch", "03:30", func(ctx context.Context) error { return fillWatcher.Check() })`; internal/scheduler/scheduler.go:269 has GetJobs but grep finds no RunNow/Trigger method and no run-job route in internal/web. Needs a new operator-gated trigger endpoint (auth surface) — a new mechanism, solve together with R-279. | 4 |
|
||||
| R-184 | P4 | FIXED-BY-LATER-WORK | Fixed by felhom.eu b55fc17d "hub v0.102.0 — refuse to vouch a version that cannot be installed (R-273)" — exactly shape (b), validate at vouch time in the hub. felhom.eu/hub/internal/web/configs.go:1358 `res := s.gitea.PackageDownloadable(ctx, t.pkg, t.version, t.file)` and :1365 `s.logger.Printf("[WARN] artifact vouch REFUSED: %s package %s is NOT downloadable (R-287)", ...)`; tag leg at :1343 Ta | 4 |
|
||||
| R-194 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Proxmox caches permissions, so a removed storage grant can still read as present for seconds to minutes; our self-repair notices only after the cache expires. | 2 |
|
||||
| R-206 | P4 | STILL-TRUE-NOT-SMALL | homelab-manifests@87dfc29 (/home/kisfenyo/git/homelab-manifests): no daemon.json template in homelab-ansible (grep finds only a comment at roles/node_housekeeping/templates/node-housekeeping.sh.j2:17 and homelab-ansible/CLAUDE.md:54). Part (b) was superseded by fc9fbb8 ("correct the expired Docker rationale"): the script now says at :13-20 do NOT add docker calls, the GC policy in daemon.json is t | 4 |
|
||||
| R-207 | P4 | FIXED-BY-LATER-WORK | Fixed by homelab-manifests fc9fbb8 "node_housekeeping: guard DRY_RUN, correct the expired Docker rationale, pin container log rotation". /home/kisfenyo/git/homelab-manifests/homelab-ansible/roles/node_housekeeping/templates/node-housekeeping.sh.j2:137 `if [[ "${DRY_RUN}" == "1" ]]; then` inside write_metrics, :138 logs "file left untouched". | 3 |
|
||||
| R-208 | P4 | STILL-TRUE-SMALL | FIX: Move the ARG VERSION/GIT_COMMIT (controller) and ARG VERSION/BUILD_TIME (hub) declarations down to just above the final `go build` RUN. — felhom-controller@7690c27 controller/Dockerfile:12 `ARG VERSION=dev` and :13 `ARG GIT_COMMIT=unknown` sit above :19 `RUN go mod download // true`; felhom.eu@53d8131b hub/Dockerfile:3 `ARG VERSION=dev`, :4 `ARG BUILD_TIME=unknown` above :9 `RUN go mod downlo | 3 |
|
||||
| R-209a | P4 | UNCHECKED | Live-only: whether DooPlex has rebooted and /var/log/felhom-store-postboot-check.log says PASS. Not read (DooPlex is Tier 2, operator ruled no reboot; this pass touches no machine). No source claim to check. | 2 |
|
||||
| R-210 | P4 | NOT-WORTH-IT | PICK close-as-accepted: 193 old controller/hub images exist only on DooPlex and cannot be re-pulled; the question is whether to delete them. | 2 |
|
||||
| R-213 | P4 | STILL-TRUE-NOT-SMALL | Row is a not-started design (live-vs-backup comparison, then put-back flow), operator-owned; nothing in source to verify against. | 1 |
|
||||
| R-230 | P4 | STILL-TRUE-NOT-SMALL | Owed rulings, not code: (a) bulk-correction ruling on MEMORY.md staleness (MEMORY.md index still carries version literals, e.g. 'ctrl 0.224.0', 'hub 0.109.0'); (c) spec-as-failing-test pilot not started. (b) closed. Operator decision required. | 2 |
|
||||
| R-246 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu@53d8131b hub/internal/store/store.go:3248 `func (s *Store) MarkEscrowStale(hostID string) error {` still has no production caller (grep: only definition + comments at offsite.go:208,216); stale_at still read (store.go:3182 clears it). Ruling owed by operator: evidential setter or retire the column (folds R-248). | 3 |
|
||||
| R-256 | P4 | STILL-TRUE-SMALL | FIX: Rewrite flash.offbox.mgr_unavailable / mgr_unreachable in both languages to say the backup service is not running yet and give a route (try again in a few minutes; if it persists, contact support). — felhom-controller@7690c27 controller/internal/i18n/locales/hu.json:1406 `"flash.offbox.mgr_unavailable": "A mentéskezelő nem elérhető.",` used at controller/internal/web/offbox_handlers.go:54 and | 4 |
|
||||
| R-261 | P4 | STILL-TRUE-SMALL | FIX: Reword the doc comment (selfbind.go:106-110) to say it is a test accessor and name the two tests that pin the auto-mint invariant (selfbind_automint_test.go, customer_delete_test.go) — or, if the operator prefers, add one post-mint production check that logs [WARN] when count != 1. — felhom.eu@53d8131b hub/internal/store/selfbind.go:111 `func (s *Store) CountSelfBindTokens(customerID string) | 3 |
|
||||
| R-263 | P4 | STILL-TRUE-SMALL | FIX: Change the comment to 'the only writer that GRANTS the role' and add a source-scanning test that finds every `.BackupTarget =` assignment in non-test settings code and fails if any other than SetBackupTarget can assign a non-false value. — felhom-controller@7690c27 controller/internal/settings/settings.go:1655 `// from every other. This is the ONLY writer of StoragePath.BackupTarget — registr | 3 |
|
||||
| R-264 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu@53d8131b scripts/wire_contract_gate.py still allowlists the six with _R264: :242 selfupdate_pending, :246 selfupdate_pending_version, :255 restore_tests.mount_parity, :258 restore_tests.mount_inventory, :281 backup.last_db_dump, :282 backup.last_integrity_check. Each reader is a design per the row. | 3 |
|
||||
| R-266 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller@7690c27 controller/internal/report/builder.go:94 `{Mount: "/", Label: "SSD", TotalGB: sysInfo.DiskTotalGB, UsedGB: sysInfo.DiskUsedGB, Percent: sysInfo.DiskPercent},` — no disk_known on the storage entry; hub has no disk_known (grep empty). Two-repo wire change gated by wire_contract_gate.py. | 3 |
|
||||
| R-279 | P4 | STILL-TRUE-NOT-SMALL | No operator/hub path to start an off-site run: grep for offbox run triggers in felhom.eu/hub/internal finds nothing; the only run entry is the customer dashboard handler (felhom-controller controller/internal/web/offbox_handlers.go:270 `if !s.backupMgr.OffboxRunnable() {`). Needs a new operator-authenticated trigger — sibling of R-177, not a duplicate (different job). | 3 |
|
||||
| R-284 | P4 | NOT-WORTH-IT | PICK close-as-accepted (close as not-a-defect): A reported 'almost full' warning on an empty disk; the code shows the warning only below 20% free and hides it by default, so the report was a reading of unrendered HTML. | 5 |
|
||||
| R-285 | P4 | STILL-TRUE-NOT-SMALL | No maintenance/expected-downtime concept in hub: `grep -rln -i 'maintenance/expected_downtime/quiet_until/snooze' felhom.eu/hub/internal` returns nothing. New mechanism (M). | 3 |
|
||||
| R-286 | P4 | STILL-TRUE-SMALL | FIX: Add one paragraph: a positive control must come from a different channel than the measurement (different query path, snapshot, API or clock); give the 2026-08-09 stale-snapshot case as the example. Put it in ONE home (pointer elsewhere). — Lesson (a) not written anywhere: grep -i 'different channel/same channel/independent channel' over documentation/runbooks/workspace-CLAUDE.md, felhom.eu/sk | 6 |
|
||||
| R-287 | P4 | FIXED-BY-LATER-WORK | Deleter established 2026-08-10 (R-267 newest-10 prune, recorded in the row itself); CI fixed by felhom-agent 53d047a "Two guards, one number: bound the published check to the retention it must live with" (R-291). felhom-agent@e06ed97 scripts/check-published-versions.py:101 `RETENTION_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "retention-policy.json")`, :213 `keep = retention_k | 5 |
|
||||
| R-288 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu@53d8131b documentation/architecture/00-capability-map.md is now 210 125 bytes / 30 937 words / 253 lines (`wc`), larger than the 134 642 bytes measured in the row; :38 still reads `*Verified 2026-07-16 against evidence corpus @ felhom.eu tip `4b18cc5``. Restructure is an M doc surgery, operator-owned. | 3 |
|
||||
| R-289 | P4 | FIXED-BY-LATER-WORK | R-182 was closed by felhom.eu ef6ac6fe (2026-08-22, register compression): documentation/backlog/CLOSED-ITEMS.md:474 `/ **R-182** / ... / **CLOSED — SHIPPED** (controller v0.194.0 + hub v0.90.0/.1, 2026-08-03) /`. The residue (digest never seen delivering) was since observed: documentation/audits/DRILL-chaos-night-2026-09-17.md:181 `backup_run_failures` „1 of 12 apps failed to back up in this nigh | 5 |
|
||||
| R-290 | P4 | STILL-TRUE-NOT-SMALL | Gate exists (felhom.eu scripts/check_stands.py) but the map itself still carries the claims: documentation/architecture/00-capability-map.md has 95 'PROVEN-LIVE' occurrences (grep -c); demoting 12 rows or writing walk documents is M and blocked on R-288 per the row. | 3 |
|
||||
| R-291 | P4 | STILL-TRUE-SMALL | FIX: Rewrite the _comment/recorded_by to cite the operator's newest-10 rule (R-267/R-287) instead of 'observed, not a ruling', and drop or correct the non-existent registry-retention.md reader. Keep the min_agent-floor note as the recorded better bound; then close R-291. — felhom-agent/scripts/retention-policy.json still says the 10 is 'NOT a ruling anyone has been able to locate' and recorded_by: | 8 |
|
||||
| R-292 | P4 | STILL-TRUE-SMALL | FIX: Make resolveArtifactSHA return a reason (not-found / unreachable / bad manual sha) and redirect to three distinct flashes (reuse artifact_unverifiable for unreachable, add artifact_version_missing, keep artifact_sha_invalid for a bad typed sha incl. the wrapper sha at :1395). — hub/internal/web/templates/configuration.html:55 still reads 'the Gitea sha lookup failed (version missing / Gitea u | 6 |
|
||||
| R-310 | P4 | STILL-TRUE-SMALL | FIX: Drop the second 'The vouched golden is' sentence when GOLDEN_CHECK_WHY already names it (or drop the version from :3060); add one runbook line: --uninstall needs an interactive terminal; --force does not bypass the typed vmid confirm. — felhom.eu/scripts/felhom-host-install.sh:3060 sets GOLDEN_CHECK_WHY="it is controller $ver, but the vouched golden is $ART_GOLDEN_VER" and :3080-3081 die "... | 6 |
|
||||
| R-315 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/scripts/wire_contract_gate.py:30-48 still documents the test as a repo-wide literal-tag search ('IT PROVES REACHABILITY OF A NAME'); ROOTS at :88 includes the R-311 escrow/retained root. No receiver-type field-by-field comparison exists. Fix requires resolving receiver mirror types — a new mechanism (M). | 5 |
|
||||
| R-325 | P4 | STILL-TRUE-SMALL | FIX: Import RETRIEVAL_STEMS from ../felhom.eu/scripts/customer_copy_vocab.py (same sibling-path pattern as controller_gates.py:48) and delete the STEMS literal; absent sibling = INCONCLUSIVE exit 2. Follow-up (felhom.eu, separate commit): remove hub_copy_gate.py's drift check, which would then fail to find STEMS. — felhom-controller/controller/scripts/retrieval_promise_gate.py:54 still has its own | 6 |
|
||||
| R-327 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/documentation/architecture/where-felhom-stands.yaml:126-130 still: id claim.code-naming, title "The same word is used for two different secrets across three surfaces; the email points at a page a rebuilt machine does not show", status: partial. Needs the operator's capability-map ruling first (dataset may not be raised on its own). | 4 |
|
||||
| R-331 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/internal/agentapi/diskverdict.go:34 uncorrectableFailCount = 64; :33 comment still defers 'growth-rate detection once the box keeps history'. No growth-rate rule found. New mechanism (M). | 4 |
|
||||
| R-336 | P4 | STILL-TRUE-NOT-SMALL | No source change reduces the ep0 poll rate (pvestatd interval is Proxmox-side, not in our repos). Design question (does the hub need a 15-min fill reading) remains; acceptance needs an ep0 access-log measurement. Scaling item, not small. | 3 |
|
||||
| R-337 | P4 | UNCHECKED | Live-only behaviour (WATCHING). From source: felhom-agent/internal/localapi/server.go:518 serves GET /backup/status and :1258 answers from s.pickLatestBackup (the in-memory store), which suggests collection cadence, but the refresh path after an out-of-schedule run was not established within the time box. | 6 |
|
||||
| R-345 | P4 | STILL-TRUE-SMALL | FIX: Delete lines 21-22 (or move them behind an explicitly named opt-in target with a comment). Whether a stale :latest already sits on the registry is a separate live check for a session allowed to query it. — felhom.eu/hub/Makefile:21 'docker tag $(IMAGE):$(VERSION) $(IMAGE):latest' and :22 'docker push $(IMAGE):latest' still present; only commit touching the Makefile is 77b5a4ce (initial). | 3 |
|
||||
| R-346 | P4 | NOT-WORTH-IT | PICK close-as-accepted (audit done, zero instances): A warning that a future reader might anchor an uptime slope on systemd's ActiveEnterTimestamp instead of the process start time. | 4 |
|
||||
| R-348 | P4 | STILL-TRUE-SMALL | FIX: Reword the comment: the backups list IS lost on restart and refills only when a backup runs; the hub's freshness VERDICT is unaffected because it looks back 7 days (hub monitor/deadline.go backupEvidenceLookback, pinned by deadline_anchor_test.go TestCheckBackupDeadlines_RestartBlindWindow_NoEvent). — felhom-agent/internal/backup/store.go:28 still reads '// Backups are unaffected — their fres | 6 |
|
||||
| R-352 | P4 | STILL-TRUE-NOT-SMALL | Placement half is an open operator ruling (SPEC-app-data-placement-2026-08-21.md, 'Viktor rules'); deploy route still has no server-side default: GetDefaultStoragePath has no caller in internal/stacks (grep returns only internal/api/router.go:1137 systemInfo). Point (2) is carried by R-368. | 4 |
|
||||
| R-364 | P4 | STILL-TRUE-SMALL | FIX: A helper that, for a pattern containing a byte >= 0x80, also runs an ASCII anchor (must hit) and a negative control (must miss) and refuses to print a zero unless both behave; documented in the felhom-evidence or ui-hungarian rule as the way to search Hungarian text. — No helper exists: ls felhom.eu/scripts shows nothing grep/accent-related, and no script mentions '0x80' or 'negative control' | 4 |
|
||||
| R-365 | P4 | STILL-TRUE-SMALL | FIX: Set AbandonOverdue when DueAt is past and render a new key ('a törlés esedékes, a következő napi karbantartáskor lefut' / English twin) instead of the future-tense sentence. — felhom-controller/controller/internal/i18n/locales/hu.json:368 'A kérésed szerint a korábbi távoli mentéseidet <strong>{{.AbandonDate}}</strong> napján véglegesen töröljük (még ... nap)'; handlers.go:1164-1167 sets Aban | 5 |
|
||||
| R-367 | P4 | NOT-WORTH-IT | PICK close-as-accepted (or delete it the next time the box is reprovisioned): One old 312 KB paperless database dump on the demo-hp test box sits under the app's old folder name; nothing reads or deletes it. | 4 |
|
||||
| R-368 | P4 | STILL-TRUE-SMALL | FIX: Reword the field comment: the deploy FORM pre-selects this path (templates/deploy.html:614); the deploy API applies no default when HDD_PATH is omitted. (The alternative — a server-side default — is a behaviour change and not small.) — felhom-controller/controller/internal/settings/settings.go:572 'IsDefault bool `json:"is_default,omitempty"` // new apps use this by default' (line moved from | 5 |
|
||||
| R-371 | P4 | NOT-WORTH-IT | PICK close-as-accepted with one line in 07-backup-architecture saying success is silent by design because failure and staleness are alarmed: The weekly off-site backup sends no 'done' event, while the two local tiers do. Failures and an 8-day staleness deadline are already alarmed. | 6 |
|
||||
| R-372 | P4 | NOT-WORTH-IT | PICK close-as-accepted: An optional idea from July: show 'this second-drive copy was never made because its source is missing' separately from 'last copy failed' in the operator screen. | 5 |
|
||||
| R-373 | P4 | FIXED-BY-LATER-WORK | Premise (20G/50G two-volume mismatch, 'nothing sets SysDataGrowGB') was retired by agent v0.120.0 one-data-volume work, commit cd6e267 'v0.120.0 — one data volume (R-165...)'. felhom-agent/internal/reconcile/bringup.go:191 '// SysDataGrowGB is a COMPATIBILITY INPUT since agent v0.120.0 (R-165). There is no longer a second' and :437 'growGB := spec.DataVolGrowGB + spec.SysDataGrowGB'; installer pas | 6 |
|
||||
| R-374 | P4 | NOT-WORTH-IT | PICK close-as-accepted, with one line in the audit saying the three are not recoverable: A July audit says three borderline cases were left unfiled but never named them. | 5 |
|
||||
| R-375 | P4 | UNCHECKED | Requires a read-only check on ep0 (token's datastore audit permission); not verifiable from source and ssh is out of scope for this checker. | 2 |
|
||||
| R-376 | P4 | STILL-TRUE-SMALL | FIX: Carry the same marker legend paragraph into the three documents written after the 2026-08-22 pass; then close the row, since 'mark as sessions touch them' is a standing practice already in the template, not a defect. — Legend present ('not yet classified') in 00..06 and 10 of documentation/architecture/, but MISSING in the newer 08-alarm-ladder.md, 09-update-architecture.md and 11-os-updates. | 5 |
|
||||
| R-377 | P4 | STILL-TRUE-SMALL | FIX: Turn each ruling's opening bold line '**S-NN — TITLE (date ...).**' into a '### S-NN — TITLE (date)' heading, changing no other byte; no compression, no reordering. — felhom.eu/CONTEXT.md:1537 '## Standing rulings' runs to EOF: 189,685 bytes, 0 '###' sub-headings, 153 bullets, 39 distinct S- ids; each ruling starts as a bold paragraph e.g. '**S-39 — "WE DO NOT KNOW" IS NEVER DRAWN AS "FINE".. | 6 |
|
||||
| R-390 | P4 | FIXED-BY-LATER-WORK | Commit 2344589a ('... runbook pveam note'); felhom.eu/documentation/runbooks/RUNBOOK-manual-build.md:154 '2. Run **`pveam update` first** — the `virgin` snapshot's template INDEX is stale too, and a stale index fails as a bogus'. | 3 |
|
||||
| R-391 | P4 | STILL-TRUE-SMALL | FIX: Take the row's second option: state in CLAUDE.md that the catalog REPORT.md carries no observations section by convention (findings go straight to the register), so gate 11 is not needed here. The runner refactor (first option) is the bigger alternative. — app-catalog-felhom.eu/scripts/catalog_gates.py has no SHARED_ / observations entry (grep 'SHARED_/observations' returns nothing); app-cata | 4 |
|
||||
| R-392 | P4 | STILL-TRUE-NOT-SMALL | ls felhom.eu/documentation/architecture shows no agent-tooling/workflow document (00-11 are all product; plus _design-review, _hub-review, _recovery-inventory). Writing a new architecture document is more than an hour and needs the operator's view of the split. | 3 |
|
||||
| R-393 | P4 | NOT-WORTH-IT | PICK close-as-accepted (superseded in practice by unprompted-work.md §2/§4): A proposed skill plus helper script to log every decision an unattended run makes. | 4 |
|
||||
| R-394 | P4 | STILL-TRUE-NOT-SMALL | wc -l felhom.eu/skills/felhom-build-deploy/SKILL.md = 186 (was 179 at filing — grew); scripts/check_skills.py:45 GRANDFATHERED still holds the exemption. Trim needs a session that can verify the build/deploy commands it keeps. | 3 |
|
||||
| R-402 | P4 | STILL-TRUE-NOT-SMALL | No hub Go/template reads last_integrity_ok/_depth (grep in hub/internal returns nothing); still allowlisted at felhom.eu/scripts/wire_contract_gate.py:163 and :220. Needs the operator's decision on what the screen says. | 3 |
|
||||
| R-416 | P4 | STILL-TRUE-SMALL | FIX: Apply the existing RULE 3 duplicate check to CLOSED-ITEMS.md too (suffixed ids like R-88a/R-88b stay distinct), and update the closed_register_gate.py:53 hole list to point at it. — Partly covered: scripts/register_shape_gate.py:120 'RULE 3 — duplicate: {rid} already has a row at line ...' (added in 462ab4a5, R-627) now refuses a duplicate id WITHIN OPEN-ITEMS.md only (REG path at :84 = OPEN- | 7 |
|
||||
| R-418 | P4 | STILL-TRUE-SMALL | FIX: Add the two missing gates to the docstring list and a test that parses the docstring's gate labels and asserts they equal [g[0] for g in GATES]. — It drifted AGAIN: felhom.eu/scripts/repo_gates.py docstring lists 1-14 (+9b) = 15 gates while GATES has 17 — 'script-tests' and 'decoy-coverage' are registered but not listed (python import: len(GATES)=17). No test compares the two. | 5 |
|
||||
| R-420 | P4 | NOT-WORTH-IT | PICK close-as-accepted (add it with the first gate that needs it): The felhom.eu gate runner cannot mark a gate as advisory-only; the controller runner can. | 3 |
|
||||
| R-421 | P4 | STILL-TRUE-NOT-SMALL | Deliberate class row ('stays open as the place the next instance is recorded'); its open instances R-422..R-426 are still open in this batch. Not a fixable item by itself. | 2 |
|
||||
| R-422 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/scripts/reuse_refs_check.py:41 PATH_RE still ends '\.(?:go/py/html/css/yml/yaml/sh)\b' — no .md. Widening it needs a false-positive walk across all four repos' REUSE.md/CLAUDE.md citations (the row says that pass is the work). | 3 |
|
||||
| R-423 | P4 | STILL-TRUE-SMALL | FIX: Discover website/**/*.html and FAIL on any page not in PAGES (or glob and keep PAGES only as exceptions); flip the decoy test to expect conviction and drop the 'site' EXEMPT entry. — felhom.eu/scripts/site_gates.py:22-27 hardcoded PAGES list; website/ today holds exactly those 9 files, so nothing is missed today, but a new page is not scanned. | 4 |
|
||||
| R-424 | P4 | NOT-WORTH-IT | PICK close-as-accepted (hole stays declared in the gate's docstring): The roadmap gate cannot tell a real defect filed as an 'idea' from a genuine idea. | 3 |
|
||||
| R-425 | P4 | STILL-TRUE-SMALL | FIX: Scan by pattern (templates/backups*.html, *offbox*.go) plus internal/i18n/locales/hu.json, or assert FILES against a discovered set so an unclassified file fails; drop its decoy-coverage exemption. — felhom-controller/controller/scripts/offbox_rename_gate.py:16-20 FILES = backups.html, offbox_handlers.go, offbox.go only; templates/backups_remote.html exists and is not scanned, and customer co | 6 |
|
||||
| R-426 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/scripts/decoy_coverage_gate.py EXEMPT now has 19 entries (loaded via python): hub-copy is gone, felhom-agent 'release-complete' is new; group (d) gates (hostinstall, wire-contract, due-checks, published, image-resolvable, volume-persistence) all still exempt. | 5 |
|
||||
| R-427 | P4 | FIXED-BY-LATER-WORK | Commit 71b8c8c6 (Backlog triage Part B: '... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS'); felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word (CLOSED, SHIPPED,'. Of the 12 named rows, R-385/387/341/378/405/88a/88b/123 are now only in CLOSED-ITEMS.md; R-190 and R-352 remain open (partly-closed, as the ro | 5 |
|
||||
| R-437 | P4 | FIXED-BY-LATER-WORK | felhom.eu 71b8c8c6 'Backlog triage Part B: 125 finished rows + 20 id-less rows moved to CLOSED-ITEMS ... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS (decoys, seen red) ... register 444 -> 325'. felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word'. Gate run today: 'closed-register gate OK — no open work fi | 4 |
|
||||
| R-445 | P4 | NOT-WORTH-IT | PICK close-as-accepted: The hub's per-app memory suggestion can be built from samples of an app that no longer runs anywhere (e.g. a 15-minute test install). | 5 |
|
||||
| R-451 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/internal/report/types.go ContainerDetailReport still carries only Name/State/CPUPercent/MemoryMB (no image field). Ruled (09 §3 decision 18), build deferred until fleet grows; needs controller payload + hub denormalisation + fleet page across two repos. | 3 |
|
||||
| R-454 | P4 | STILL-TRUE-SMALL | FIX: Add a gofmt -l gate (fails on any listed file, INCONCLUSIVE if gofmt missing) to controller_gates.py, see it red on today's tree, then one gofmt -w formatting commit for the 12 files. — The five files named are now clean (gofmt -l controller/internal/web/ prints nothing), but `gofmt -l controller` in felhom-controller lists 12 OTHER files today: cmd/controller/main.go, internal/agentapi/diskv | 5 |
|
||||
| R-457 | P4 | STILL-TRUE-SMALL | FIX: Read the six candidates; for each date literal that feeds an assertion evaluated against time.Now(), derive it from now (as the R-457 fix did). The faked-future-date CI instrument is a separate, larger idea and should be split out or dropped. — No later commit references R-457 beyond the filing release (felhom-controller 38d28b5 v0.234.0 / 998aa31 REPORT). No faked-clock CI run exists (grep f | 4 |
|
||||
| R-460 | P4 | NOT-WORTH-IT | PICK close-as-accepted: BookStack's uploaded files cannot be checked automatically after an upgrade; only its database can. | 2 |
|
||||
| R-464 | P4 | FIXED-BY-LATER-WORK | Lesson homed and harness uses the correct probe. app-catalog-felhom.eu b7ef0c4 'upgrade-test.py: record the engine's own view of its datadir'; app-catalog-felhom.eu/scripts/upgrade-test.py:278 '"mariadb-upgrade --check-if-upgrade-is-needed --user=root "'. felhom.eu d6837d98 (SPIKE R-459); felhom.eu/documentation/architecture/09-update-architecture.md:1647 '1. **Ask the engine, not the log.** Maria | 4 |
|
||||
| R-488 | P4 | UNCHECKED | The claim is a measured suite runtime (5.5 min); confirming it needs running go test ./internal/backup, which I did not run (read-only; backup tests may reach real docker on DooPlex). No commit after filing (felhom-controller 24d7c54) mentions R-488 or a test-speed change in internal/backup (git log --grep on internal/backup since 2026-09-13: empty), so it is likely still true. | 3 |
|
||||
| R-492 | P4 | STILL-TRUE-SMALL | FIX: Remove Paths.HDDPath, its env binding and each reader's dead global branch, keeping the per-app/discovered fallbacks each reader already uses. — cfg.Paths.HDDPath still defined and read: controller/internal/config/config.go:167 'HDDPath string `yaml:"hdd_path"`', :453 envStr("FELHOM_PATHS_HDD_PATH", &cfg.Paths.HDDPath); readers internal/report/builder.go:69, internal/monitor/healthchec | 4 |
|
||||
| R-494 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/hub/internal/cloudflare/ holds only unblock.go; no tunnel/DNS creation code in hub (grep cfd_tunnel: none). Building it is a new Cloudflare-API mechanism on the hub; operator ruled it non-blocking. | 3 |
|
||||
| R-501 | P4 | FIXED-BY-LATER-WORK | felhom.eu a4993272 'CLAUDE.md: the CI-check recipe was wrong in two ways, both measured today'. felhom.eu/CLAUDE.md:176 'rows — a run can sit several pages earlier. **Scan every page** and match on `head_sha`; with a'; recipe at CLAUDE.md:166-168 loops every page. | 3 |
|
||||
| R-502 | P4 | STILL-TRUE-SMALL | FIX: Register bootstrap-modes.sh in repo_gates.py behind a docker-available check that reports INCONCLUSIVE (never pass) when docker/the felhom-iso-assistant image is absent, plus a decoy (a broken banner must turn it red). — felhom.eu/scripts/iso/test/bootstrap-modes.sh exists; `grep -rn 'bootstrap-modes/bootstrap_modes' scripts/*.py .gitea/workflows` in felhom.eu returns nothing — no gate or CI | 3 |
|
||||
| R-503 | P4 | NOT-WORTH-IT | PICK close-as-accepted (as DECLINED by ruling; the three measurements stay in the closed row for any future reversal): An idea, offered and not chosen: the installer would pick the disk itself when there is exactly one. | 2 |
|
||||
| R-504 | P4 | UNCHECKED | The claim (iso.felhom.eu/ returns 404) is live-only; I may not curl hosts. felhom.eu/documentation/runbooks/VOLUNTEER-first-hour.md:14 still says '`iso.felhom.eu/` itself still has no index — R-504'. Fix needs a Cloudflare rule the operator owns. Cosmetic; households use felhom.eu/letoltes (website/letoltes.html exists). | 3 |
|
||||
| R-507 | P4 | STILL-TRUE-NOT-SMALL | Needs measuring QEMU input-send-event or a VNC client against a live VM on felhom-pve — a live-machine spike, not a source change. No later commit references R-507. | 2 |
|
||||
| R-525 | P4 | STILL-TRUE-NOT-SMALL | Row itself states it is a new unmeasured mechanism (forwardAuth / Quantum proxy auth) needing a scratch-guest spike. | 1 |
|
||||
| R-526 | P4 | STILL-TRUE-NOT-SMALL | felhom.eu/hub/internal/tenantsync/client.go:110 '// Deprovision DESTROYS the customer's PBS namespace, all its backup groups, and its token — the'; no token-only op exists. Needs a new op on protected ep0 and an operator yes/no. | 3 |
|
||||
| R-527 | P4 | NOT-WORTH-IT | PICK close-as-accepted (with the corrected facts: flag read into LockedFields, enforced only by uncalled UpdateStackConfig): A catalog flag that marks some settings 'locked after install' changes nothing visible: the page makes every setting read-only anyway, and the edit path that would honour the flag is never called. | 8 |
|
||||
| R-532 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Vaultwarden's web page shows a sign-up form even though sign-ups are off; the server then refuses it. | 3 |
|
||||
| R-541 | P4 | STILL-TRUE-NOT-SMALL | Row: needs a new copy/re-key/release mechanism and a design; no later commit references R-541. Far off per re-rank (0.3% full, one pool box). | 2 |
|
||||
| R-544 | P4 | STILL-TRUE-SMALL | FIX: Change the log line to state the effect, e.g. 'host deleted: %s (escrow custody demoted to retained)' when escrow existed, and drop the boolean name from the text. — felhom.eu/hub/internal/web/hosts.go:917 's.logger.Printf("[INFO] host deleted: %s (escrow deleted: %v)", hostID, deleteEscrow)'. | 3 |
|
||||
| R-551 | P4 | NOT-WORTH-IT | PICK close-as-accepted (add 'walk R-546 readiness branches' to the next fresh-install checklist instead): The escrow 'waiting for the agent' screens are proven by tests but never seen on a real box in that state. | 2 |
|
||||
| R-555 | P4 | STILL-TRUE-SMALL | FIX: Strip Go // and /* */ comments and template {{/* */}} before TOKEN_RE in receiver_tokens; add a decoy 'tag named only in a receiver comment must convict'. Newly surfacing tags each become a finding (allowlist with reason or a row). — felhom.eu/scripts/wire_contract_gate.py:451 'def receiver_tokens(repo_root):' tokenises whole files: line 482 'toks.update(TOKEN_RE.findall(fh.read()))' — no com | 3 |
|
||||
| R-564 | P4 | STILL-TRUE-SMALL | FIX: Add split-form Hungarian patterns (állíthatók? vissza, (hoz/szerez/nyit)\w* vissza), register the Hungarian occurrences found (the seven already reviewed in English), and add a planted split-verb decoy. — felhom-controller/controller/scripts/retrieval_promise_gate.py:54 'STEMS = ["visszaállíthat", "visszaszerezhet", "visszahozhat", "visszanyit"]' — joined forms only, no split-verb pattern. | 3 |
|
||||
| R-567 | P4 | STILL-TRUE-SMALL | FIX: Add (eq .Page "storage_init") (eq .Page "storage_attach") to $storageOpen and mark the Meghajtók link active for them. — controller/internal/web/templates/layout.html:83 '{{$storageOpen := or (eq .Page "storage") (eq .Page "storage-network")}}' — storage_init/storage_attach not included; storage_handlers.go:351 'data := s.baseData(tmpl, title)' passes the template name as Page. | 4 |
|
||||
| R-568 | P4 | STILL-TRUE-SMALL | FIX: Sort rows by diskKey(d) (durable id, falling back to name) before returning. — controller/internal/web/disk_health.go:124-152 diskHealthRows appends rows in resp.Disks order; no sort in the file (grep 'sort.' in disk_health.go: none). | 3 |
|
||||
| R-569 | P4 | STILL-TRUE-SMALL | FIX: Add KindErrorf-style sentinels in internal/stacks (protected, not found, not deployed/still running, not orphaned), a statusFor helper per handler family replacing the three Contains blocks. — controller/internal/api/router.go:742 'if strings.Contains(err.Error(), "protected") {', also :745, :1018, :1021, :1024 ('not deployed'/'still running'), :1102, :1105, :1108 ('not orphaned'). | 3 |
|
||||
| R-570 | P4 | STILL-TRUE-NOT-SMALL | Fallback still present: controller/internal/web/handlers.go:1050 'offboxStaleWarningMarker = "nincs mentésre jelölt alkalmazás"'; producer internal/backup/offbox.go:1168. Closing depends on a fleet condition (every box one off-site run on >=0.251.0) — a watch, not a fix. | 3 |
|
||||
| R-571 | P4 | STILL-TRUE-SMALL | FIX: Add a short section to 07 listing the six failure classes, what each means for the customer, and that restic/ssh signatures are external; add an alert-placement paragraph (inline under storage bars vs top banner) to 02. — grep ClassifyOffsiteFailure/PageOnly/Inline in felhom.eu/documentation/architecture/07-backup-architecture.md and 02-controller-module-map.md: no hit. Classifier lives at fe | 3 |
|
||||
| R-574 | P4 | STILL-TRUE-NOT-SMALL | controller/internal/web/handler_debug.go still carries 40 lines with accented Hungarian string literals (grep -cP count); last touched by 0c702f8 v0.279.0, not converted. Labelling ~40 literals page-copy vs payload, adding en/hu keys and parity fixtures exceeds an hour. | 3 |
|
||||
| R-576 | P4 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/scripts/i18n_go_parity.py has no call-site argument-count or '+'-adjacency check (grep verb/argument: only VERB_RE/strip_verbs for text equality, lines 76-78, 196-199). Parsing multi-line Go call arguments reliably from Python with decoys is likely >1 h. | 4 |
|
||||
| R-577 | P4 | STILL-TRUE-NOT-SMALL | Waiting on an operator decision (what the share feature promises a stranger); felhom.eu/documentation/architecture/10-localisation.md table row 'the two guest share pages, the catch-all / a stranger / nobody / **no globe** / — (R-577, the operator's)'. | 2 |
|
||||
| R-579 | P4 | STILL-TRUE-NOT-SMALL | Gate deliberately deferred until R-554 deletes the first-boot wizard; R-554 is still OPEN (OPEN-ITEMS.md:131). Versionless links remain in controller/internal/setup/templates/setup_*.html:8 '<link rel="stylesheet" href="/static/style.css">' (8 files). | 5 |
|
||||
| R-588 | P4 | STILL-TRUE-SMALL | FIX: Name the single home documentation/tests/iso-release-<ver>-<date>/ in the Result-recording section, and add a pointer dir/README for 1.28.0 to its audit record (evidence-backup-promise-2026-09-16/phaseD-iso-gate.txt). Optional existence check left out. — felhom.eu/documentation/runbooks/iso-release-gate.md:319-322 'Result recording' says only 'in the release report' — names no home. documenta | 4 |
|
||||
| R-591 | P4 | STILL-TRUE-SMALL | FIX: Deep-copy Meta.I18n in deepCopyStack (map plus nested values). — The copy is deepCopyStack in controller/internal/stacks/manager.go (row says Copy()); it deep-copies AppConfig (:1137-1148), DeployFields (:1162), OptionalConfig (:1174), Integrations (:1186) and has no I18n line (grep I18n in manager.go: none). Meta.I18n defined at internal/stacks/metadata.go:94. | 4 |
|
||||
| R-594 | P4 | STILL-TRUE-SMALL | FIX: Add ALLOWLIST_EN of (app, path, reason); registered occurrences pass, unregistered convict, and any entry matching nothing is itself a failure. — app-catalog-felhom.eu/scripts/check-copy-i18n.py: grep ALLOWLIST/allowlist — no hit; no way to register a true occurrence. | 3 |
|
||||
| R-599 | P4 | STILL-TRUE-SMALL | FIX: Make both 409 bodies say when the last report arrived and when deletion opens (last report + configured stale threshold), and name the wait in target-selection.md's drill section. — felhom.eu/hub/internal/web/hosts.go:898 'http.Error(w, "Host is ONLINE — deletion is refused (a live agent would receive 401s permanently).", http.StatusConflict)' — no last-report age or opening time. target-sele | 4 |
|
||||
| R-602 | P4 | FIXED-BY-LATER-WORK | felhom.eu e02bc038 'hub v0.119.0 — ... R-596/R-598 closed' added the finding; felhom.eu/documentation/architecture/10-localisation.md:809-812 'the `felhom_lang` cookie and got the **Hungarian** page for `en`. ... The cookie is the right instrument for the anonymous claim page and the **wrong**' and :509 '`langFor`'s order is fixed: `?lang=` → **the household's setting when a session exists**'. Onl | 4 |
|
||||
| R-603 | P4 | STILL-TRUE-SMALL | FIX: Add a test helper that compares against html.EscapeString(want) and use it in the render tests that assert English copy; the bundle gate with a 27-entry allowlist is the larger alternative. — No gate or helper: grep for html.EscapeString(want/'/R-603 in felhom-controller/controller scripts+internal: none. controller/internal/i18n/locales/en.json has 27 lines containing an apostrophe today | 4 |
|
||||
| R-605 | P4 | STILL-TRUE-SMALL | FIX: Give harness-level refusal its own exit code (e.g. 3 = REFUSED) in both scripts and map it to a distinct label in catalog_gates.py. — app-catalog-felhom.eu/scripts/catalog_gates.py:122 'VERDICT = {0: "OK", 1: "FAILED", 2: "INCONCLUSIVE"}' — harness refusal and per-app undetermined both exit 2 and print the same word. | 3 |
|
||||
| R-610 | P4 | NOT-WORTH-IT | PICK close-as-accepted: A power cut landing inside the sub-second `starting` phase has never been measured; all three live cuts landed in `verifying`, which runs the same recovery code. | 5 |
|
||||
| R-617 | P4 | FIXED-BY-LATER-WORK | felhom.eu@462ab4a5 (2026-09-22) documentation/architecture/09-update-architecture.md:1969 "`POST /api/v1/repos/migrate` is the route that works (the project's Gitea tokens carry" - continues at :1970 "`write:repository` but not `write:user`, so `POST /user/repos` answers 403"; recipe at :1979. The one-line note the row asked for exists (in the architecture doc rather than operations/). The optiona | 4 |
|
||||
| R-618 | P4 | STILL-TRUE-NOT-SMALL | Remaining open work = the controller-side idea (let `verifying` accept docker's own `healthy`). grep for docker health status in felhom-controller@7690c27 controller/internal/stacks/update.go returns nothing; no R-618 reference in Go source. It is an undecided design question (operator), not a defect; the three probe fixes (app-catalog@793c4fb) and the gate scripts/check-probe-matches-compose.py a | 4 |
|
||||
| R-619 | P4 | STILL-TRUE-SMALL | FIX: In getDeployFields, copy meta.DeployFields and set Required=true for every field with Type=="password" before writing the response (copy, do not mutate the shared metadata; the web deploy page uses GetDeployFields separately and is untouched). — felhom-controller@7690c27 controller/internal/api/router.go:395 `meta, appCfg, err := r.stackMgr.GetDeployFields(name)` then :402 `"metadata": meta | 6 |
|
||||
| R-621 | P4 | STILL-TRUE-NOT-SMALL | Capture is done: felhom-controller@7690c27 controller/internal/stacks/update.go:1054 `outDir := filepath.Join(dir, "hold-logs", ts)`. The open part (show it on the app page's hold panel / logs fallback) is not built: grep for hold-logs/holdLogs in controller templates and handlers returns only update.go:1050-1082 and undo.go:535,550 (writers). Surfacing needs a page change with HU/EN copy and a de | 4 |
|
||||
| R-624 | P4 | STILL-TRUE-NOT-SMALL | Row's own latest update: remaining class is vaultwarden (closed sign-up by design) and code-server; the open decision is whether the harness may hold an app's admin secret (operator). Not verifiable further from source; needs a decision, not a fix. | 2 |
|
||||
| R-644 | P4 | UNCHECKED | About the live state of gokapi on scratch guest 9202 (crash-loop, deployed:true). Only the box shows it; no ssh allowed. | 1 |
|
||||
| R-652 | P4 | STILL-TRUE-NOT-SMALL | app-catalog@917a779 templates/romm/.felhom.yml:249 still carries `"memory_peak_pct": 80.9` with `"memory_tight": true` and no `memory_basis: anon` (contrast paperless-ngx/.felhom.yml:249 `"memory_basis": "anon"`). Needs a live re-measure of romm plus an undecided cache-thrash rule. | 4 |
|
||||
| R-654 | P4 | NOT-WORTH-IT | PICK close-as-accepted: opengist 1.15 moved its pages under /-/; an old /login bookmark answers 404. The front page redirects correctly. | 3 |
|
||||
| R-687 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Three live proofs a scratch box cannot give (a 3-hour leg, a failing off-site leg, a files_may_change step without a whole copy) plus one log text that names the window's deadline instead of a manually started leg's deadline. | 5 |
|
||||
| R-688 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Deleting a customer does not remove their Cloudflare tunnel and DNS records; the dialog now says so and lists what to remove by hand. | 4 |
|
||||
| R-691 | P4 | STILL-TRUE-NOT-SMALL | Open work = the Tier 2 (second-drive) path of Use/Load is not live-proven; needs a two-drive Tier-0 box (9202 has one drive). Live-only gap, not a source defect; not checkable from source. | 2 |
|
||||
| R-693 | P4 | STILL-TRUE-NOT-SMALL | grep for memory_scales_with_limit / scales_with_limit across app-catalog-felhom.eu, felhom-controller/controller and felhom.eu/scripts returns nothing - no basis that tells growth-to-fill from pressure exists. Needs a design (new harness signal). | 3 |
|
||||
| R-705 | P4 | FIXED-BY-LATER-WORK | The remaining half (manual whole-guest backup) EXISTS and predates the row: felhom-controller bbed5af (v0.47.0, 2026-06-12) 'backups page — whole-guest backup visibility + manual trigger'. Live source @7690c27: controller/internal/web/backup_handlers.go:324 `case r.URL.Path == "/api/guest-backup/trigger" && r.Method == http.MethodPost:`; :340 `if err := s.backupTrigger.TriggerNow(); err != nil {`; | 8 |
|
||||
| R-707 | P4 | STILL-TRUE-NOT-SMALL | Open work = live proof that the gate OPENS for seerr, outline and rallly (needs a media server / e-mail on a test box). Live-only proof gap; the gating itself is in source (catalog 6faf432 per row). | 2 |
|
||||
| R-718 | P4 | STILL-TRUE-SMALL | FIX: Add a key app_info.close_signup_restart ("The app restarts once for this." / HU twin) and render it under the close card when the app has an after_setup env switch (the same SignupNative fact); add the same sentence to the gate-open confirmation where after_setup.env exists. — felhom-controller@7690c27 controller/internal/web/templates/app_info.html:112 `<p>{{T "app_info.close_signup_text"}}< | 6 |
|
||||
| R-719 | P4 | STILL-TRUE-NOT-SMALL | Built: felhom.eu hub/internal/web/selfbind.go:160 "// R-719 (v0.126.0): „Új linket kérek" on an expired or used link." Open part is the operator's review of the changed shape and a live mint+send proof (unit-proven only) - an operator decision, not a CC fix. | 3 |
|
||||
| R-725 | P4 | STILL-TRUE-SMALL | FIX: Reword bind.invalid.body to point at the button below (e.g. 'Ha lejárt, kérj újat lent.' / 'If it has expired, ask for a new one below.'), keeping the operator alternative; optionally drop the ✔ glyph the console font renders as 'V' and update the golden. — felhom.eu hub/internal/i18n/locales/hu.json:79 `"bind.invalid.body": "A hivatkozás 7 napig érvényes. Ha lejárt, kérj újat az ügyfélszolgá | 6 |
|
||||
| R-731 | P4 | STILL-TRUE-NOT-SMALL | The shape-switch control lives only in audit tools: felhom.eu/documentation/audits/catalog-currency-2026-09-30/00-currency.py, 04-analyse.py; no standing currency script in app-catalog-felhom.eu/scripts or felhom.eu/scripts (ls/grep for currency/shape returns only golden_currency_gate.py, which is unrelated). Making it standing means promoting a registry-reading tool with tests - more than an hour | 4 |
|
||||
| R-734 | P4 | STILL-TRUE-NOT-SMALL | grep for '.immich' / hash ignore list in app-catalog-felhom.eu/scripts/*.py and templates/immich/.felhom.yml returns nothing - no exclusion exists. The row says the rule change needs an operator word; calibre-web shows the mark is sometimes right, so the rule needs design. | 3 |
|
||||
| R-739 | P4 | STILL-TRUE-NOT-SMALL | app-catalog@917a779 templates/wanderer/docker-compose.yml:117 `image: getmeili/meilisearch:v1.36.0`; grep MEILI_UPGRADE_DB in the compose returns nothing. Remaining: the template switch, a fixture (PocketBase create refused) and a measured step on the bench - live work. | 3 |
|
||||
| R-759 | P4 | STILL-TRUE-NOT-SMALL | Five checklist rows of wger need live measurement on 9202 (2.5, 3.7, 6.3, 8.2, 9.1); not verifiable from source. | 2 |
|
||||
| R-760 | P4 | STILL-TRUE-SMALL | FIX: Read the vikunja 2.6.0 image config for a HEALTHCHECK/shell; if none and the image has no shell, add a comment saying why there is no compose healthcheck (like adventurelog-frontend's R-655 comment); otherwise add a healthcheck of the family the image supports (REUSE.md §2). No image: line moves, so no catalog_since. — app-catalog@917a779 templates/vikunja/docker-compose.yml: service vikunja | 4 |
|
||||
| R-761 | P4 | STILL-TRUE-SMALL | FIX: Change the comment to name `{slug}-logo.svg` (preferred) and `{slug}-logo.png` (fallback), matching config.go AppLogoURL/AppLogoPNGURL; comment-only. — app-catalog@917a779 templates/paperless-ngx/.felhom.yml:22 `# Logo: {assets.base_url}/assets/{slug}-logo.webp` vs felhom-controller controller/internal/config/config.go:511 `return fmt.Sprintf("/static/assets/%s-logo.svg", slug)` and | 3 |
|
||||
| R-764 | P4 | STILL-TRUE-NOT-SMALL | grep smtp/mail in app-catalog templates/wger/.felhom.yml and docker-compose.yml finds only first_steps text (.felhom.yml:74 'Add meg az email címedet a beállításokban'); no smtp_mapping. A mapping needs a live boot proof with mail off (REUSE.md §2) - more than an hour; wger is hidden. | 3 |
|
||||
| R-766 | P4 | FIXED-BY-LATER-WORK | Hub releases after the assets push (felhom.eu 40f07429, 2026-10-01): hub v0.131.0 (2026-10-04) .. v0.136.0 (d4be9f6f, 2026-10-05). The build copies website assets: felhom.eu scripts/build-hub.sh:98 `cp "${WEBSITE_ASSETS_DIR}"/*-logo.svg "${BUILD_DIR}/assets/" 2>/dev/null // true`, and the hub build workspace /mnt/5_hdd/felhom.eu/build/felhom-hub/workspace/assets/ holds radicale-logo.svg + 3 screen | 8 |
|
||||
| R-768 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Grimoire is not built because upstream rules out public exposure and ships no image for v1.x; the row only watches for that to change. | 1 |
|
||||
| R-769 | P4 | STILL-TRUE-NOT-SMALL | New-app idea waiting on an operator decision (a fork as new upstream, after R-767). No source to check. | 1 |
|
||||
| R-770 | P4 | STILL-TRUE-NOT-SMALL | New-app idea waiting on the operator's go/no-go (CC recommends not building). No source to check. | 1 |
|
||||
| R-771 | P4 | STILL-TRUE-NOT-SMALL | New-app idea waiting on the operator's go/no-go. No source to check. | 1 |
|
||||
| R-779 | P4 | STILL-TRUE-NOT-SMALL | Live proof gap on the real Cloudflare tunnel needing the operator's phone off wifi; not checkable from source. | 1 |
|
||||
| R-781 | P4 | STILL-TRUE-SMALL | FIX: After the clone, delete the real onboarding/<app>.md records (all but _TEMPLATE.md and the exempt wger.md the cases use) from the scratch clone, so genuine cases judge only the records they build; alternative: point the stand-in sibling at the real felhom.eu documentation tree read-only. — app-catalog@917a779 scripts/test_gate_decoys.py:498 `sh(["git", "clone", "-q", "file://" + ROOT, cat], c | 6 |
|
||||
| R-786 | P4 | STILL-TRUE-NOT-SMALL | app-catalog@917a779 onboarding/sparkyfitness.md has 8 '/ open' rows, incl. :18 0.5, :20 0.7, :28 1.6, :29 1.7, :58 5.4, :68 8.3 (plus 0.1 licence = R-784, 2.1 = R-807). Most need live measurement (runtime internet, phone sign-in, second memory watch). | 3 |
|
||||
| R-793 | P4 | NOT-WORTH-IT | PICK close-as-accepted: Four apps ship enterprise/BUSL code that is off as Felhom runs them; the row reminds us never to enable EE features or the -enterprise meilisearch image. | 2 |
|
||||
| R-794 | P4 | STILL-TRUE-NOT-SMALL | app-catalog@917a779 seven redis 7 images: dawarich/docker-compose.yml:164 `image: redis:7.4-alpine`, docmost:86, immich:123, outline:88, nextcloud:103, paperless-ngx:125, romm:136 `image: redis:7-alpine`. Moving each needs a harness-proven ladder step (7 apps). | 3 |
|
||||
| R-796 | P4 | NOT-WORTH-IT | PICK close-as-accepted: MeTube's browser/phone 'send to MeTube' helpers cannot pass the family gate; households paste links in the page. | 2 |
|
||||
| R-797 | P4 | NOT-WORTH-IT | PICK close-as-accepted: CI's single-repo clone cannot check rule 3 of the family-gate gate; it says NOT CHECKED, and the pre-push hook checks it. | 2 |
|
||||
| R-798 | P4 | STILL-TRUE-SMALL | FIX: Remove the dead SWAGGER_ENABLED line (or rename to API_DOCS_ENABLED=false, which v3.5.0 reads and defaults to false). No image: line moves. — app-catalog@917a779 templates/grimmory/docker-compose.yml:29 ` - SWAGGER_ENABLED=false` still present. | 2 |
|
||||
| R-799 | P4 | STILL-TRUE-SMALL | FIX: Add "download_type": "video" to the POST /add body. — app-catalog@917a779 scripts/upgrade_fixtures_box.py:2183 `data=json.dumps({"url": self.URL, "quality": "best", "format": "any", "auto_start": True}), method="POST")` - no download_type. | 2 |
|
||||
| R-804 | P4 | NOT-WORTH-IT | PICK close-as-accepted: plant-it's image repository does not exist; the template is already abandoned and not installable, and no box runs it. | 2 |
|
||||
| R-805 | P4 | STILL-TRUE-NOT-SMALL | app-catalog-felhom.eu scripts/check-volume-persistence.py:342 'if m["class"] == "named-declared" and m.get("files", 0) == 0:' — only named volumes judged empty; binds not. Last change 917a779 (R-788). The rule change itself is small, but it flips Grimmory/komga/paperless-ngx/radarr/sonarr to UNDETERMINED and needs a live re-sweep to regenerate the verdict tables; the row also asks for a decision. | 6 |
|
||||
| R-806 | P4 | STILL-TRUE-SMALL | FIX: Make routed_ports return the traefik loadbalancer.server.scheme (reuse upgrade_boxport LB_SCHEME_RE) and have the GET exercise use that scheme with curl -k for https. The gramps-web :5000 non-answer stays a separate live look (narrow the row to it). — app-catalog-felhom.eu scripts/check-volume-persistence.py:586 'code = _sh(a + [f"http://{ip}:{port}{path}"], timeout=40)' — plain http always; | 6 |
|
||||
| R-807 | P4 | STILL-TRUE-NOT-SMALL | Per-app upload seeds for 13 apps (claper, crafty-controller, dawarich, docmost, gramps-web, immich, outline, sparkyfitness, tandoor, vikunja, wger, wishlist, zipline) + plex/wanderer; each needs a live fixture run. Gate rule at scripts/check-volume-persistence.py:342 still makes empty declared volumes UNDETERMINED (917a779). | 3 |
|
||||
| R-814 | P4 | UNCHECKED | Live Hetzner console state (box 611421 status); not visible in source. Operator action only. | 2 |
|
||||
| R-815 | P4 | UNCHECKED | First GC completion on felhom-offsite is PBS server-side live state; grep of documentation found no GC completion record (DIAG-backup-missed-2026-07-26.md:43 'prune/GC history NOT COLLECTED'). | 3 |
|
||||
| R-816 | P4 | STILL-TRUE-NOT-SMALL | Needs a live exercise of six failure classes on a scratch guest; no source change can close it. | 2 |
|
||||
| R-817 | P4 | STILL-TRUE-SMALL | FIX: Add a dated correction under decision 56: the swap rolls back to the image running when the swap began (controllerswap.go Swap/rollback); the kept previous image is for a hand roll-back. Do not rewrite the ruling itself. — felhom.eu documentation/architecture/09-update-architecture.md:619 '56. **A box keeps the controller image it runs and the one before it** (the self-update's roll-back targ | 8 |
|
||||
| R-818 | P4 | STILL-TRUE-SMALL | FIX: Add a dated correction note under the v0.109.0 entry (and the hub CHANGELOG mentions at :695/:701) naming the real closed rows from CLOSED-ITEMS.md. Also present in felhom-controller/CHANGELOG.md:3107 and :3234 (R-330/R-331 for v0.224.0/v0.225.0) — a second repo; either note it there too or narrow the row. — felhom.eu hub/CHANGELOG.md:759 '## v0.109.0 — the Backup card told every operator tha | 6 |
|
||||
| R-819 | P4 | STILL-TRUE-SMALL | FIX: Let rule 3 accept an id found in CLOSED-ITEMS.md (and check the stand's status agrees), fix the R-273/R-356 dangling ids, register the gate in repo_gates.py with a decoy. — Ran python3 scripts/check_stands.py (read-only): still convicts e.g. 'fail.stolen-machine: register id R-281 is not in OPEN-ITEMS.md', 'fail.customer-self-restore: register id R-356 ...'. grep 'stands' in scripts/repo_gate | 6 |
|
||||
| R-832 | P4 | STILL-TRUE-NOT-SMALL | Deferred roadmap item: a third-location copy of ep0 is money + operator decision (decision 71). Nothing in source to change. | 1 |
|
||||
| R-844 | P4 | STILL-TRUE-NOT-SMALL | Needs a household timeline on the controller, which does not exist (row: 'when the box gets a household timeline'). A new surface, not a fix. | 2 |
|
||||
| R-855 | P4 | STILL-TRUE-SMALL | FIX: Add a small helper (e.g. osSvc.DockerNightsEffective()) mapping negative→0 and 0→2, and print that in the start log. — felhom.eu hub/cmd/hub/main.go:450 'logger.Printf("[INFO] osupdates: the Docker engine set is approved only by the operator, after %d healthy ring-0 night(s)", osSvc.DockerNights)' prints the raw value; internal/osupdates/service.go:169 'negative means none'. | 4 |
|
||||
| R-856 | P4 | STILL-TRUE-NOT-SMALL | Row is marked an operator design question (crash-restart suppression for app mails); not a defect yet. | 1 |
|
||||
| R-857 | P4 | STILL-TRUE-SMALL | FIX: Have newest_baked accept an optional suffix after the date and, for equal versions, prefer the newest bake-log timestamp (or refuse two dirs for one version); add 're-vouch at once after a same-version re-bake' to the runbook. — felhom.eu scripts/golden_currency_gate.py:146 'EVIDENCE_RE = re.compile(r"^golden-(\d+)\.(\d+)\.(\d+)-\d{4}-\d{2}-\d{2}$")' and :237-238 'found.sort() / return found[ | 7 |
|
||||
| R-878 | P4 | STILL-TRUE-NOT-SMALL | Next action is 'measure a large volume first' — a live measurement; the fix direction is a behaviour change to the catch-up. | 2 |
|
||||
| R-881 | P4 | STILL-TRUE-SMALL | FIX: Add an rm -f of /usr/local/sbin/felhom-priv-apply to the uninstall step (tolerate-absent, like felhom-pbs-apply at :1161) and fix the :1679 comment; ships at the next installer tag. — felhom.eu scripts/felhom-host-install.sh: grep 'felhom-priv-apply' → no hit anywhere in the installer (uninstall does not remove it); :1679 '+ guest-hook snippet under /var/lib/vz/snippets/ (agent-installed at r | 4 |
|
||||
| R-884 | P4 | UNCHECKED | Live ArgoCD diff on DooPlex (forbidden to touch here). Related: homelab-manifests mon-system/monitoring.yaml:399-426 is the same prometheus Deployment R-211 concerns. | 2 |
|
||||
| R-885 | P4 | STILL-TRUE-SMALL | FIX: Finish and push the in-progress script_tests_gate.py (walks scripts/ for test_*.py, exit-code verdict, nesting guard) registered in repo_gates.py; coordinate with the session that owns the dirty tree. — On main (b018ca90) scripts/repo_gates.py runs no test_*.py suite. NOTE: the felhom.eu working tree holds UNCOMMITTED work for exactly this row by another session: '?? scripts/script_tests_gate | 6 |
|
||||
| R-30 | P3 | STILL-TRUE-NOT-SMALL | Design change (presence from the Dir-2 long-poll instead of the report clock), size M; no commit with R-30 after aa9c08f0 (filing). | 3 |
|
||||
| R-31 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu hub/internal/web/configs.go:1594 'd, err := s.offsite.ProvisionOffsite(ctx, cfg.CustomerID, in)' still in-request; :1584 detaches from the request context (mid-cancel fixed) but no async/status card. | 5 |
|
||||
| R-35 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller controller/internal/report/config_refresh.go:65 'config-refresh: applied config_version=%d — self-restarting to load it'; sessions are in-memory only: controller/internal/web/auth.go:259-260 's.sessions[token] = &session{'. Hot-apply or persisted sessions is a design change with security weight. | 6 |
|
||||
| R-49 | P3 | STILL-TRUE-NOT-SMALL | app-catalog-felhom.eu templates/immich/.felhom.yml:28-34 backup block has no cache/volume exclusion; row itself says a capture-set exclusion needs its own ruling (data-loss-shaped). | 4 |
|
||||
| R-50b | P3 | FIXED-BY-LATER-WORK | Claim 'fetched via fetch_raw from raw/branch/main — no tag, no pin' no longer true: bee68484 (installer v1.23.0, R-110/R-183) pinned fetch_raw to the vouched agent tag — felhom.eu scripts/felhom-host-install.sh:533 '"$GITEA_BASE/$GITEA_OWNER/$AGENT_REPO/raw/tag/v$ART_AGENT_VER/$path" \' (leg b). Leg (c)-like signed delivery: felhom-agent c9fa2e7 (R-840 config bundle) and configs/test_felhom_config | 7 |
|
||||
| R-78 | P3 | STILL-TRUE-NOT-SMALL | An owed operator decision + spike (local_api authority); not a code defect. | 1 |
|
||||
| R-79 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller controller/internal/monitor/healthcheck.go:100 'fmt.Sprintf("SSD disk usage critical: %.0f%%"', :213 'Protected container not running: %s'; rendered raw at controller/internal/web/alerts.go:241 'Message: issue, // ON THE WIRE ... not ours to translate; slice 3'. Whole-surface, seam needs a spike. | 5 |
|
||||
| R-118 | P3 | STILL-TRUE-SMALL | FIX: Only statfs the mount when s.devicePresent(d.MountPath) is true (else leave capacity zero/unknown); put statfsCapacity behind a seam var for the test. — felhom-agent internal/localapi/disks.go:401 'if total, used, okc := statfsCapacity(d.MountPath); okc {' — no device-presence guard on the union path; devicePresent exists (disks.go:985) and is used just above (:381) for BoundUnderParent. | 6 |
|
||||
| R-121 | P3 | FIXED-BY-LATER-WORK | 3d7a2761 (hub v0.135.0, R-530/R-604 'boxes left behind listed and alarmed'): felhom.eu hub/internal/osupdates/service.go:79 'EventAgentBehind = "agent_behind" // warning, operator' with :180 'AgentBehindAfter: a box runs an agent older than the vouched one this long → an operator alarm' (7 d window, the staleness window the row asked for). | 5 |
|
||||
| R-126 | P3 | STILL-TRUE-SMALL | FIX: Skip IsNetwork() paths in the export-destination list and refuse them in isValidDrivePath for the export POST (keep scanning for import if wanted, via a separate list). — felhom-controller controller/internal/web/handler_export.go:377-386 storageDriveList() appends every s.settings.GetStoragePaths() entry with no IsNetwork() filter; the predicate exists at controller/internal/settings/setting | 6 |
|
||||
| R-127 | P3 | STILL-TRUE-NOT-SMALL | Leg (a) still true: grep 'data_key: true' in app-catalog-felhom.eu templates → only adventurelog, dawarich, homebox, papra, sparkyfitness; n8n N8N_ENCRYPTION_KEY, wanderer POCKETBASE_ENCRYPTION_KEY, calcom CALENDSO_ENCRYPTION_KEY, bookstack APP_KEY unflagged (templates/n8n/.felhom.yml:38 etc.). Leg (b) (regenerated DB password vs restored PGDATA) needs a design choice. Leg (a) alone is a ~45-min c | 6 |
|
||||
| R-130 | P3 | STILL-TRUE-SMALL | FIX: Take the cheap honest branch: rename to RECOMMENDED_MIN_LVM_GIB and reword the warning to 'below the recommended …' (making it refuse would change install behaviour and needs a ruling). — felhom.eu scripts/felhom-host-install.sh:348 'HARD_MIN_LVM_GIB=120 # a useful appliance won't fit below this on local-lvm' and :1760 '... // log_warn "local-lvm free ~${free_gib} GiB < hard min ${HARD_MIN_ | 4 |
|
||||
| R-132 | P3 | UNCHECKED | Whether HUB_PW was rotated is out-of-band operator state; not visible in source. | 1 |
|
||||
| R-136 | P3 | STILL-TRUE-SMALL | FIX: Introduce a const sessionCookieName = "__Host-hub_session" and use it at all five sites (Path=/, Secure, no Domain already hold). Every operator logs in once more; plain-HTTP browser access stops (Basic auth unaffected). — felhom.eu hub/internal/web/server.go:857 'Name: "hub_session",' and readers at server.go:806, :896, :918 and apps.go:351. | 4 |
|
||||
| R-137 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller controller/internal/cloudflare/waf.go:18 'globalRuleDesc = "[felhom-geo] Global"', :21 'appRuleDescPrefix = "[felhom-geo] app:"' — still not namespaced. Two-repo M change. | 3 |
|
||||
| R-138 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller controller/internal/infra/infra.go:157-159 writes CF_DNS_API_TOKEN when d.CFAPIToken != ""; no shared-zone guard in hub (grep shared.zone: none). Needs a policy decision first; no shared zone exists today. | 4 |
|
||||
| R-179 | P3 | STILL-TRUE-SMALL | FIX: In uninstall step 4c, before the umount loop: stop+disable every mnt-felhom\x2ddrives-*.automount/.mount unit, rm their files from /etc/systemd/system, daemon-reload (tolerate-absent; still never umount -l/-f). — felhom.eu scripts/felhom-host-install.sh uninstall section :1121-1157 handles felhom-shared-parent and umounts under /mnt/felhom-drives, but grep 'x2ddrives/automount' → no hit: the | 6 |
|
||||
| R-180 | P3 | STILL-TRUE-SMALL | FIX: In the same pre-flight block, die (byo) / die (appliance) if ARCHIVE_STORAGE is not in PVE_STORAGES, with a message naming --acl-storages. — felhom.eu scripts/felhom-host-install.sh:1800 'if pvesm status --storage "$ARCHIVE_STORAGE" ...' checks existence only; PVE_STORAGES=(local local-lvm felhom-pbs) at :322; no ARCHIVE_STORAGE ∈ PVE_STORAGES assertion. | 5 |
|
||||
| R-190 | P3 | NOT-WORTH-IT | PICK close-as-accepted: A storage permission vanished once on demo-felhom in August and nobody knows why. Since agent 0.124.1 the agent puts it back by itself and mails the operator when it happens. | 4 |
|
||||
| R-200 | P3 | FIXED-BY-LATER-WORK | The remaining half (customer-facing recovery-code form: yell → R form → preview) shipped as the recovery screen: felhom-controller 636c51e 'R-193: the recovery screen — unlocking, and only unlocking (v0.200.0)'; controller/internal/web/templates/recovery.html:82 '<form id="unlock-form" method="POST" action="/recovery/unlock" autocomplete="off">', routed at internal/web/server.go:602. Plumbing half | 8 |
|
||||
| R-211 | P3 | STILL-TRUE-NOT-SMALL | homelab-manifests (/home/kisfenyo/git/homelab-manifests @87dfc29) mon-system/monitoring.yaml:420 'image: prom/prometheus:v3.15.0', :426 '--web.enable-lifecycle'; grep 'reload/checksum/config' → none. The manifest edit is small, but it rolls the production Prometheus on DooPlex (operator territory) and the same Deployment is OutOfSync per R-884 — do the two together. | 6 |
|
||||
| R-231 | P3 | STILL-TRUE-NOT-SMALL | Owner operator; DooPlex /opt/backup/scripts remains host state. Partially touched by cea8502f (scripts/hub-db-backup versioned in felhom.eu, cites R-231) but that covers only the hub-DB push, not /opt/backup/scripts or the same-disk/no-off-site facts. | 4 |
|
||||
| R-235 | P3 | FIXED-BY-LATER-WORK | felhom.eu c033b3b6 'ISO 1.28.0 source: the console stops showing the pairing code once bound (R-535)'. scripts/iso/felhom-bootstrap.sh:538: `print_bound_banner # R-535: replace the pairing code on the console with the truth`. Same defect already CLOSED twice in CLOSED-ITEMS.md as R-535 (line 232) and R-214 (line 200, 'proven on a fresh install'). | 4 |
|
||||
| R-240 | P3 | STILL-TRUE-SMALL | FIX: Replace the producer string at offbox.go:1168 with wording that drops 'Sikeres' but keeps the lowercase marker substring, e.g. 'Ez a futás semmit nem mentett: nincs mentésre jelölt alkalmazás'; update the tests that pin the literal. — felhom-controller/controller/internal/backup/offbox.go:1168: `warns = append(warns, "Sikeres — nincs mentésre jelölt alkalmazás")`; web/handlers.go:1068-1069 re | 8 |
|
||||
| R-242 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/scripts/golden_currency_gate.py:23-24: 'It does **NOT** check that the golden was **VOUCHED**, because the vouched version lives ONLY in the hub's `hub_settings` table'; :40 'That vouch half is STILL open after 2026-09-13'. Last gate commits 5ef0f52b/ae59c31a did not add a vouch check. | 5 |
|
||||
| R-244 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/hub: no cascade leg touches app_log_issues — only writers are internal/store/telemetry.go:176-209 (upsert), :537 `DELETE FROM app_log_issues WHERE last_seen < ?`, :556/:575 operator deletes by app/id. cmd/hub/main.go:1004: `if n, err := s.PruneStaleIssues(time.Now().Add(-30 * 24 * time.Hour))` (since a757bee0, hub v0.4.0). | 7 |
|
||||
| R-250 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/hub/internal/offsite/offsite.go:71-72: `var defaultScanBackoff = []time.Duration{2 * time.Second, 4 * time.Second, 8 * time.Second, 16 * time.Second, 30 * time.Second}`; scanner.go:40 dials plain "tcp" (no A-record preference); no R-250 commit. | 7 |
|
||||
| R-251 | P3 | STILL-TRUE-SMALL | FIX: In offsiteNewestPerTag skip the 'felhom-offbox' marker tag (move the constant into package backup and reuse it from web); consider whether '_shares' should render as a named row or be skipped. — felhom-controller/controller/internal/backup/offbox_inventory.go:102-108: loops `for _, tag := range sn.Tags` and adds every non-empty tag to `newest` — no filter for 'felhom-offbox'. The marker filte | 9 |
|
||||
| R-255 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller: still no page-wide runtime secret-sentinel test — `grep -l sentinel internal/web/*_test.go` hits only edge_safe_status_test.go, i18n_parity_test.go, recovery_test.go; controller/scripts/secret_in_markup_gate.py remains the only all-template net. No commit cites R-255. | 7 |
|
||||
| R-257 | P3 | STILL-TRUE-SMALL | FIX: Rewrite flash.offbox.not_orphaned (hu + en) to say what the customer tried, that it does not apply now, and where to look, without 'offsite'/'elárvult', e.g. 'A távoli mentés rendben van, nincs mit félretenni. Ha gondod van vele, írj nekünk.'; wording sign-off from the operator (owner Viktor). — felhom-controller/controller/internal/i18n/locales/hu.json:1409: `"flash.offbox.not_orphaned": "Az | 5 |
|
||||
| R-262 | P3 | STILL-TRUE-SMALL | FIX: One-repo fix: narrow the comment to say hostBackup is field-for-field and hostRestoreTest is a deliberate SUBSET (lists mount_parity/mount_inventory as not modelled), and add a hub test that names the two agent fields as known-unmodelled so a future addition must edit it. Adding the fields + fixture is a two-repo change (byte-identical golden) and stays a separate choice. — felhom.eu/hub/inte | 7 |
|
||||
| R-269 | P3 | STILL-TRUE-SMALL | FIX: On a map hit, also stat the store and reload when its size differs from loadedSize before answering (one cheap stat per auth), so a rotated-out token is rejected without depending on an unrelated miss. — felhom-agent/internal/localapi/tokenstore.go:173-176: `if vmid, ok := s.byHash[want]; ok { if subtle.ConstantTimeCompare(...) == 1 { return vmid, true } }` — a superseded token's hash is stil | 6 |
|
||||
| R-270 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/internal/bootstrap/bootstrap.go:255: `if cfg == nil // cfg.LocalAPI.Endpoint != "" {` (fill-only, never refreshes); DetectEndpointDrift (bootstrap.go:369-399) compares only the endpoint. No R-270 commit. | 5 |
|
||||
| R-271 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/internal/channelhealth/checker.go:152: `if prev != "" && prev != "up" {` (unseeded->up is silent); checker.go:87 alert text still says '(re-bootstrap)'. No R-271 commit. | 4 |
|
||||
| R-274 | P3 | FIXED-BY-LATER-WORK | felhom.eu eb600872 'R-297: installer compares a local golden against the manifest before using it'. scripts/felhom-host-install.sh:3074: `if golden_local_matches_manifest "$GOLDEN_VOLID"; then` (digest vs ART_GOLDEN_SHA, else baked marker vs ART_GOLDEN_VER; otherwise ignores the local golden and fetches, or dies if the operator named it, :3080). Line numbers differ from the triage note (2855-2905) | 5 |
|
||||
| R-275 | P3 | STILL-TRUE-SMALL | FIX: Purge every sibling `"${agent_cfg}".*` (and then rmdir/rm the config dir) instead of `.bak*`; fix the WIPED line wording. — felhom.eu/scripts/felhom-host-install.sh:1059: `for _cfgbak in "${agent_cfg}".bak*; do [[ -e "$_cfgbak" ]] && run rm -f "$_cfgbak"; done` — glob still misses agent.json.campaign8-before, .campaign9-prev, .pre-e-target-move, .pre-prunegate.bak; :814 still claims 'config ( | 6 |
|
||||
| R-276 | P3 | STILL-TRUE-SMALL | FIX: In run_uninstall (full scope): stop+disable wg-quick@wg-felhom and remove /etc/wireguard/wg-felhom.conf via run(); add it to WIPED, and add a KEPT line 'the hub-side WireGuard peer registration — remove it in the operator UI'. — felhom.eu/scripts/felhom-host-install.sh: no reference to wg-felhom/wg-quick anywhere (grep 'wg-quick\/wg-felhom' empty); _uninstall_statement (:807-845) lists neithe | 6 |
|
||||
| R-277 | P3 | STILL-TRUE-SMALL | FIX: For UsageStr use fmtBytesAuto when the usage is below 1 GB (keep GB for the quota and the bar), so a non-empty repo never renders as 0.0 GB. — Part (a) FIXED by felhom.eu f5c9411e 'R-331 (hub half): the Backup card reads `offsite`, not the dead `backup` fields (v0.109.0)' (hub/internal/web/backup_card.go uses fmtBytesAuto :135-142). Part (b) still true: hub/internal/web/offsite_box.go:54 `ret | 8 |
|
||||
| R-282 | P3 | FIXED-BY-LATER-WORK | felhom.eu 4d6ec7c 'hub v0.104.0: ... the hub half of the naming (R-295)' + controller v0.211.0 (R-295 CLOSED, CLOSED-ITEMS.md:207) + R-323 hub v0.105.0. hub/internal/notify/templates.go:204: '// R-295, HUB HALF (2026-08-13). ONE NAME PER SECRET, and it is „Beállító kód".'; hub/internal/claim/engine.go:51 `EmailReenroll EmailKind = "reenroll"` (mail names the setup page a rebuilt box shows). | 5 |
|
||||
| R-283 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/hub/internal/claim/engine.go:7: '// engine: a rotation bumps the generation (single active code) and NEVER clears claimed_at.'; ReissueForReenroll (engine.go:195-212) rotates and mails but leaves the claim set — the hub still shows the customer as claimed after a guest rebuild. No R-283 commit. | 5 |
|
||||
| R-298 | P3 | UNCHECKED | The template gate is still there: felhom-controller/controller/internal/web/templates/storage.html:364 `if(d.role==='user-data'){` else protected (:368). BUT the agent no longer reclassifies a backup-target drive: felhom-agent/internal/localapi/disks.go:1230-1236 ('WHY THIS IS NOT A ROLE RECLASSIFICATION ... the drive that now holds the whole-guest archives is ALSO the enrolled user-data drive') a | 10 |
|
||||
| R-306 | P3 | STILL-TRUE-SMALL | FIX: Make _state_put (and _state_mark) return 0 when PREFLIGHT_ONLY is true, so a preflight-only run writes nothing; the real run's preflight records ownership again. — felhom.eu/scripts/felhom-host-install.sh:418: `$DRY_RUN && return 0` (only DRY_RUN short-circuits _state_put); :1851/:1854 `_state_put dnsmasq_preexisting yes/no` run unguarded in preflight, while :226 says '--preflight-only: ... n | 6 |
|
||||
| R-314 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller: StopAbandon is called only from cmd/controller/main.go:244 (CLI) — `grep -rn StopAbandon` shows internal/backup/offbox_abandon.go:374 and that CLI call, no web handler. | 4 |
|
||||
| R-317 | P3 | STILL-TRUE-SMALL | FIX: Probe the dnsmasq unit (e.g. /usr/lib/systemd/system/dnsmasq.service or /lib/systemd/system/dnsmasq.service) behind a small stat seam instead of /usr/sbin/dnsmasq. — felhom-agent/internal/lanresolver/lanresolver.go:107: `if _, err := os.Stat("/usr/sbin/dnsmasq"); err != nil { // metadata read, no privilege needed` — still probes the dnsmasq-base file. No R-317 commit. | 4 |
|
||||
| R-330 | P3 | STILL-TRUE-NOT-SMALL | felhom-agent/internal/hub/report.go:406-425 SmartSummary carries reallocated/pending/offline_uncorrectable + NVMe set only — no 187/188/199 fields. Wire change across agent + hub (+ controller), declared M. | 3 |
|
||||
| R-332 | P3 | STILL-TRUE-NOT-SMALL | Closing condition is live-only (a real degrading disk or an injection through agent /disks -> controller -> hub). Last related commits ea16a21b/2fa1efc narrowed the restart half only; no commit records a live Hiba-from-counters verdict. | 3 |
|
||||
| R-333 | P3 | STILL-TRUE-NOT-SMALL | (b) felhom-agent/internal/storage/hostops.go:374: `out, stderr, err := h.runner.Run(ctx, h.bins.Smartctl, "-a", "-j", device)` — no -n standby. (a) still an operator decision (Viktor decides). | 5 |
|
||||
| R-338 | P3 | UNCHECKED | felhom.eu/documentation/operations/nodes.md:86-88 now says demo-hp 'Agent config shape (R-50 island): local_api on 169.254.253.1:8443/vmbr9, guest eth1 169.254.253.2/30', and :84 records demo-hp was reprovisioned (address 192.168.0.87 -> 192.168.0.104, read 2026-09-21). Whether the reprovisioned box is actually on the island is a live-box fact (agent.json, pct config) not provable from source. | 5 |
|
||||
| R-340 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/scripts/felhom-tenantsync.sh: no health op and no 8007 probe (grep '8007\/health)' empty). Needs an ep0 (protected) script version bump + hub signal (M). | 3 |
|
||||
| R-349 | P3 | STILL-TRUE-NOT-SMALL | felhom-agent/internal/hub/report.go:282-292 reports only WrapperSHA256; no agent binary sha field in the report (grep AgentSHA256 finds only the hub's manifest entry, hub/internal/api/handler.go:2726). R-349 commits 40d857b/910fd911 are the manual correction only. | 4 |
|
||||
| R-350 | P3 | UNCHECKED | Whether the hub operator password was rotated after 2026-08-20 lives only in the hub DB / operator; git log shows no rotation record (only 910fd911 filing it). Not determinable from source. | 3 |
|
||||
| R-362 | P3 | STILL-TRUE-SMALL | FIX: When MkdirAll/write of the restore destination fails with EACCES/ENOENT, check whether the destination's drive is disconnected/decommissioned or no longer a mountpoint, and return a Hungarian error naming the drive instead of the raw permission error. — felhom-controller/controller/internal/backup/offbox_restore.go:380, offbox.go:1929, shares_restore.go:116: `return fmt.Errorf("restore dir: % | 6 |
|
||||
| R-363 | P3 | STILL-TRUE-SMALL | FIX: Replace sched.Daily("fill-watch", "03:30", …) with sched.Every("fill-watch", time.Hour, …) (scheduler.go:104), keep the startup check. — felhom-controller/controller/cmd/controller/main.go:1546: `sched.Daily("fill-watch", "03:30", func(ctx context.Context) error { return fillWatcher.Check() })`. Watcher emits on escalation only with a persisted band (internal/fillwatch/fillwatch.go:133, :231) | 6 |
|
||||
| R-388 | P3 | STILL-TRUE-NOT-SMALL | Product direction, operator's call; recorded as [DESIGN — DIRECTION] in documentation/architecture/08-alarm-ladder.md §8 per the row. Nothing in source to fix. | 2 |
|
||||
| R-401 | P3 | STILL-TRUE-NOT-SMALL | Event-triggered watch row: felhom-controller/controller/internal/backup/offbox_integrity.go:63 `const integrityCheckTimeout = 30 * time.Minute`, :78 `var integritySlowNoticeThreshold = 5 * time.Minute` — unchanged; the trigger (slow WARN on a large store) has not been recorded. | 3 |
|
||||
| R-409 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller/controller/internal/backup/recovery_unit.go:58 `Checksums map[string]string json:"checksums" // sha256 of captured compose/ files`; writers at :147, :152, :202 hash only compose/.felhom.yml/app.yaml — no db_dumps or volume_dumps hash. | 4 |
|
||||
| R-412 | P3 | NOT-WORTH-IT | PICK close-as-accepted: A narrow race: if a recovery unit is destroyed inside an off-site run after its own dump leg, the push ships the just-rebuilt hollow unit; the next run repairs it and the WARN line now says it carried no data. | 4 |
|
||||
| R-433 | P3 | STILL-TRUE-NOT-SMALL | Hetzner answered (ticket per the row): file-level snapshot access is a MAIN-account capability. hub/internal/hetznerapi/hetznerapi.go still has no snapshot read method (only size_snapshots usage, :83-87). The owed proof (read a file from a snapshot with the main account; forced-command append-only key) is a live, credential-bound operator act. | 4 |
|
||||
| R-435 | P3 | STILL-TRUE-NOT-SMALL | felhom.eu/hub/internal/monitor/offsite.go:255: `snapshotDropFraction = 0.5 // more than half the history gone in one step` — no per-tag second signal exists. | 4 |
|
||||
| R-440 | P3 | STILL-TRUE-NOT-SMALL | app-catalog-felhom.eu @917a779: 15 templates still have no `update_ladder:` in .felhom.yml — bentopdf code-server glance gokapi gramps-web homebox homepage jellyfin onlyoffice plant-it plex recipe-importer seerr vaultwarden wanderer (calibre-web got its first step in 53a4a1d). For these, pins float with no recorded digest. | 8 |
|
||||
| R-444 | P3 | STILL-TRUE-NOT-SMALL | No fstrim anywhere in felhom-agent or felhom.eu/scripts (grep 'fstrim' over *.go/*.sh empty). Needs a new periodic host job (agent, privileged, sudoers/bundle) and possibly an operator surface. | 3 |
|
||||
| R-446 | P3 | DUPLICATE | of R-440 — felhom-controller/controller/internal/stacks/updateorder.go:96: `if len(s.CatalogDigests) == 0 // s.CatalogTestedAt.IsZero() { return false }` — blind only for apps with no ladder entry, i.e. the same 15 templates R-440 lists (app-catalog has no update_ladder for them). Both rows close by the same act: each app's first proven ladder step (R-462). | 6 |
|
||||
| R-450 | P3 | FIXED-BY-LATER-WORK | The row's only remainder was 'the other ten PostgreSQL apps need two-venue proof'. R-463 CLOSED 2026-09-30 by felhom.eu 25cb3eb9 ('The last six PostgreSQL apps decided'): 8 of 11 moved by the box's own conversion, 3 (zipline, adventurelog, immich) stay by decision 42; CLOSED-ITEMS.md:223. Source proof: app-catalog templates/docmost/docker-compose.yml:62 'image: postgres:18-alpine' (also rallly:67, | 8 |
|
||||
| R-458 | P3 | NOT-WORTH-IT | PICK close-as-accepted: A frozen (pinned-behind) app can receive a newer health check from .felhom.yml. The only result is a false 'degraded'/dead-app alarm, never data loss, and only for type: api probes with an expect block. | 12 |
|
||||
| R-462 | P3 | STILL-TRUE-NOT-SMALL | Ongoing multi-session work (fixtures and ladders per app). Last progress: catalog e6f3ec2 (2026-09-30), audits/more-night-apps-2026-09-30/. 21 apps still have no ladder (per the row). Not checkable as done from source. | 3 |
|
||||
| R-468 | P3 | STILL-TRUE-NOT-SMALL | A standing pre-customer arrangement, not a defect. The mechanism is live: felhom.eu/scripts/golden_currency_gate.py:137 reads documentation/tests/golden-waiver.yml. The waiver file is currently ABSENT: deleted in felhom.eu 5efe6dae (2026-10-04, 'golden 0.292.0 vouched ... golden waiver deleted'). The row retires only at the first external install, so it stays as a watch. | 4 |
|
||||
| R-469 | P3 | STILL-TRUE-SMALL | FIX: Reword the rule heading at CLAUDE.md:103 and the 'What is NOT lifted' paragraph (:112-114) to say that PostgreSQL crosses a major one app at a time with engine_conversion + both-venue proof (decision 35) and that MySQL stays refused. Then close R-469 citing R-463's closure. Do not delete the gate: it is now the per-app enforcement. — What stood between this row and its close was R-463, now CL | 8 |
|
||||
| R-489 | P3 | FIXED-BY-LATER-WORK | The residual (a unit-restore-recreated volume has no compose label, so the remove answered []) was fixed in felhom-controller 206b035 (v0.268.0, R-658). controller/internal/stacks/delete.go:1089-1090: '// appVolumeSet is every volume the removal accounts for: the ones carrying the project label AND the // ones the app's definition declares that Docker holds by name (R-658, v0.268.0).' delete.go:70 | 5 |
|
||||
| R-498 | P3 | STILL-TRUE-SMALL | FIX: When building the app info view, rewrite each first_steps entry: replace '<word>.DOMAIN' with the stack's real address (installed SUBDOMAIN + customer domain when installed, else the template default subdomain + customer domain). Use the same approach as known_login.go:67. — Still literal: grep -l '\.DOMAIN' over app-catalog templates/*/.felhom.yml -> 58 of 58; e.g. templates/bookstack/.felho | 8 |
|
||||
| R-516 | P3 | STILL-TRUE-NOT-SMALL | By its own text it now waits for a Hungarian walk on a box with a second drive (items 4, 7, 8, 9, 10) and needs a separate row for item 11. That is live-box work, not source. No commit after 2026-09-20 names R-516 as closed. | 3 |
|
||||
| R-521 | P3 | STILL-TRUE-NOT-SMALL | No storage-disconnect suppression of app_start_failed: controller/cmd/controller/main.go:2796 'down := (stacks.IsDownState(st.State) // crashLooping) && !userStopped && !quiesced[st.Name]' (only user-stop/quiesce suppress), and controller/internal/notify/notifier.go:718 emits app_start_failed per newly-down app. It also needs the hub cooldown semantics changed (per-key cooldown outliving storage_r | 6 |
|
||||
| R-522 | P3 | STILL-TRUE-NOT-SMALL | No tunnel-connection signal in the controller: grep for TunnelConnected/cloudflared connection state in controller/internal -> none. The tile comes from container metadata: controller/internal/web/inframeta.go:21-22 '"cloudflared": { DisplayName: "Cloudflare Tunnel"'. Fixing it needs a new state source (cloudflared metrics or the hub-push result) wired into the tile, plus a live internet-cut valid | 5 |
|
||||
| R-531 | P3 | STILL-TRUE-NOT-SMALL | Both measurements are done (audits/evidence-drill-0243-2026-09-16/). What remains is an operator design question about the crash-loop budget (a slow loop every 20 min is never paused). That is an operator decision, not code. | 2 |
|
||||
| R-540 | P3 | STILL-TRUE-NOT-SMALL | Still a single pool box: felhom.eu/hub/cmd/hub/main.go:367 'poolBoxID, _ := strconv.ParseInt(os.Getenv("HETZNER_POOL_BOX_ID"), 10, 64)'. It needs a selection-rule design and eventually a second box (money). No risk today (0.3 % full per the row). | 3 |
|
||||
| R-542 | P3 | UNCHECKED | The controller passes the agent's 'initialize' list through untouched: controller/internal/web/agent_disk_handlers.go:160-162 mergeAttachCandidates only appends to Attach. The agent puts every candidate under initialize: felhom-agent internal/localapi/disks.go:438 'initialize = append(initialize, c) // every unclaimed disk can be initialized'. Whether a REGISTERED in-guest drive still counts as 'u | 8 |
|
||||
| R-545 | P3 | STILL-TRUE-NOT-SMALL | Still no un-configure route: controller/internal/web/server.go:744-783 lists /backup/offbox/{config,toggle,enable-all,offer-dismiss,run,reset,status,restore,place,reconstitute,verify-copy/delete,confirm-escrow,inject-password}; nothing removes a target. The fix is a new destructive customer action (shred data/offbox/) with a hub-escrow refusal check (R-241 rule), Hungarian/English copy and a UI. T | 4 |
|
||||
| R-547 | P3 | STILL-TRUE-SMALL | FIX: Also schedule the fill-watch on an interval (e.g. sched.Every("fill-watch-fast", 15*time.Minute, ...) calling the same fillWatcher.Check) next to the daily run, and log the cadence. Alternative if the operator prefers: state in 08-alarm-ladder.md that a transient full disk is out of scope. — Still daily: controller/cmd/controller/main.go:1546 'sched.Daily("fill-watch", "03:30", func(ctx conte | 8 |
|
||||
| R-548 | P3 | STILL-TRUE-NOT-SMALL | The row's original fix shape SHIPPED: felhom-agent 0722b2c (2026-09-24, R-685) checks before start, internal/backup/runner.go:279 '... so a new one needs about %s (old archives are removed only after a successful backup)'. Shown in the UI by controller 44ae4de (v0.272.0). The 2026-09-30 addendum is still true and is the open part: the local tier is refused for ever (10 refusals on demo-hp) because | 6 |
|
||||
| R-552 | P3 | STILL-TRUE-SMALL | FIX: Add Manager.ClearInterruptedRestore(stack) (delete from opInterrupted + persistRestoreRecordLocked under m.mu). Call it in removeStack next to ClearUpdateHold, and log when it cleared something. — The interrupted-restore notice is cleared only at controller/internal/backup/opstatus.go:61 'delete(m.opInterrupted, stack)' inside BeginRestoreOp. removeStack (controller/internal/api/router.go:955 | 6 |
|
||||
| R-554 | P3 | STILL-TRUE-NOT-SMALL | Still present: controller/internal/setup/ (setup.go, handlers.go, csrf.go, network.go, templates/), and controller/cmd/controller/main.go:328-330 'if setup.NeedsSetup(cfg) { ... runSetupMode(cfg, logger)'. The fix deletes a package and adds a new waiting page (HU+EN copy) with a red-proof. It also needs a check of drill/golden reliance on .needs-setup (controller/internal/web/handler_debug.go refe | 5 |
|
||||
| R-562 | P3 | STILL-TRUE-NOT-SMALL | Needs an operator word on the Hungarian number/date format (the row says so) and a deliberate Hungarian-byte change release with parity re-capture. Not checked further. | 2 |
|
||||
| R-565 | P3 | STILL-TRUE-SMALL | FIX: Add an ASCII Hungarian word regex to TestI18nEnglishPages (seeded from i18n_extract.py ASCII_HU plus the words releases B/C found: mp, db, FIGYELEM, jelenlegi, majd a(z), Konfig, Megtartva, helyi, Befejezve, automatikus, kedd/szerda/szombat, szint), applied after the data mask. — The English page test detects only accented letters: controller/internal/web/i18n_parity_test.go:563 'func huLette | 6 |
|
||||
| R-573 | P3 | FIXED-BY-LATER-WORK | felhom-controller 7c4a33b (v0.258.0, 'the last four Hungarian things an English household met ... R-573 the two channel banners'). controller/internal/web/alerts.go:113 'func (am *AlertManager) SetAgentChannelAlert(down bool, msgKey, msg string) {' and :136 'func (am *AlertManager) SetEndpointDriftAlert(drift bool, msgKey, msg string) {'. Both set MessageKey with msg only as a fail-open fallback. | 4 |
|
||||
| R-575 | P3 | STILL-TRUE-SMALL | FIX: Make memoryVerdict return (refusal error, warningKey string, warningArgs []any) instead of a msgHU string, and render the warning at the deploy answer with the request's language (the Alert/UpdateRefusal pattern). — Still a plain string: controller/internal/stacks/deploy.go:1456 'func (m *Manager) memoryVerdict(newReqMB, newLimitMB, releasedReqMB, releasedLimitMB int) (refusal error, warning | 5 |
|
||||
| R-578 | P3 | STILL-TRUE-NOT-SMALL | The lock-reentrancy guard is still one test in one package: controller/internal/backup/offsite_diag_test.go:190 'TestNoteHelpersAreNotCalledUnderTheSettingsLock'. No other package has one, and there is no gate (grep for R-578 in controller -> none). The fix needs a cross-package AST gate that knows which methods read settings (or a re-entrant read path in Settings). That is a new mechanism, likely | 5 |
|
||||
| R-581 | P3 | STILL-TRUE-SMALL | FIX: In GetCustomers, join on MAX(id) per customer_id instead of MAX(received_at), and add ', id DESC' to the ORDER BY at store.go:1393 and :1446. — Still no tie-break: felhom.eu/hub/internal/store/store.go:1329 'SELECT customer_id, MAX(received_at) as max_time' joined on 'r.received_at = latest.max_time' (:1333). A same-second tie returns BOTH rows (a duplicate customer in GetCustomers), not just | 6 |
|
||||
| R-584 | P3 | NOT-WORTH-IT | PICK close-as-accepted: Helper scripts with the shared DEMO controller password inline were left in a demo guest's /tmp. The cleanup rule exists, but no mechanism enforces it. | 5 |
|
||||
| R-585 | P3 | STILL-TRUE-NOT-SMALL | Still finished Hungarian from callers: controller/internal/notify/notifier.go:408 'n.PushEvent("backup_failed", "error", message, BackupDetails{Error: errMsg})', :487 offbox_enlarge_blocked, :497 db_dump_failed. None of the six types is in convertedProducers (controller/internal/notify/message_customer_test.go:39). The hub still has no customerMessages entry for offbox_enlarge_blocked (felhom.eu/h | 5 |
|
||||
| R-586 | P3 | NOT-WORTH-IT | PICK close-as-accepted: The ISO bootstrap harness runs only at each ISO release gate (G16), not on every push. | 6 |
|
||||
| R-587 | P3 | STILL-TRUE-SMALL | FIX: At the start of build-felhom-iso.sh, refuse (non-zero exit, named reason) when any *.rootpw.txt exists in the output dir. In the SKILL publish block, add a pre-check line that aborts the rclone copy if a *.rootpw.txt is present in the publish source. — The files are gone (ls /mnt/5_hdd/felhom.eu/felhom-iso/out / grep -c rootpw -> 0). The guard is still not built: the publish still relies on t | 6 |
|
||||
| R-593 | P3 | STILL-TRUE-SMALL | FIX: Move 'Az alkalmazás aldomainje' to SUBDOMAIN, give AUTH_SECRET its own description (e.g. 'A munkamenetek aláírásához használt kulcs — ne generáld újra'), add the two English i18n.en descriptions, and re-capture papra's entries in copy_freeze/hu.json with the reason in the commit. — Still wrong: app-catalog-felhom.eu/templates/papra/.felhom.yml:47 ' description: "Az alkalmazás aldomainje"' | 5 |
|
||||
| R-600 | P3 | STILL-TRUE-SMALL | FIX: Call the wgsync reconciler's Trigger() before the COMPLETE log line (nil-safe), and make the line say 'wg peer removal pushed' or 'queued for the next wgsync push' depending on whether a syncer is wired. — Log line unchanged: felhom.eu/hub/internal/web/customer_delete.go:310 's.logger.Printf("[INFO] customer DELETE cascade COMPLETE for %s (journal #%d) — full teardown", customerID, journalID) | 6 |
|
||||
| R-607 | P3 | UNCHECKED | The row asks first for a live reproduction loop (push a tag, sync, read catalog_images on a timer) and has no diagnosis. Why the sync reports 'nincs változás' while the cache moved, and when CatalogImages refreshes, are live-box behaviour I could not settle from source in the time box. No commit after d19f07ea (filing) names R-607 as fixed. | 6 |
|
||||
| R-612 | P3 | STILL-TRUE-NOT-SMALL | The memory half is fixed (catalog a5a729a, 'wishlist 512M (R-612)'). The open half, making a failed first-boot seed visible, needs a new detection mechanism (read the seed's exit/log, or a probe that checks the Role/Group rows). No commit addresses it. | 4 |
|
||||
| R-613 | P3 | STILL-TRUE-NOT-SMALL | uptime-kuma is fixed (catalog a5a729a). The open half is a sweep of all 58 templates for probes that pass on a setup wizard. That needs per-app live inspection, so it is not small. No commit names it. | 3 |
|
||||
| R-615 | P3 | STILL-TRUE-SMALL | FIX: Before the fetch, run 'git remote set-url origin <buildRepoURL()>' (or read origin and re-clone on mismatch), and log at INFO, masked, when the remote changed. The appended drill-folder residue (stack folders the live catalog lacks) is a separate drill-teardown item and is not part of this fix. — Still inert: controller/internal/sync/sync.go:279 clones only 'if _, err := os.Stat(gitDir); os.I | 7 |
|
||||
| R-616 | P3 | STILL-TRUE-SMALL | FIX: Clone and fetch with the credential-free RepoURL, and supply credentials per command via '-c http.extraHeader=Authorization: Basic <b64>' (or GIT_ASKPASS env) only when username+token are set. Pairs naturally with the R-615 set-url fix (set-url to the bare URL). — Still true: controller/internal/sync/sync.go:327-331 buildRepoURL injects 'https://%s:%s@' and the clone at :283-288 passes that U | 5 |
|
||||
| R-622 | P3 | FIXED-BY-LATER-WORK | adventurelog v0.13.0 was diagnosed, fixed and promoted with a two-venue test record in app-catalog-felhom.eu 06ea7da (2026-09-27, 'adventurelog: v0.12.1 -> v0.13.0 with its health and world-data fixes in the same commit (R-655, 09 decision 41)'; bench healthy in 217 s, box 9202 through the guarded Update in 204 s). templates/adventurelog/docker-compose.yml:13 ' image: ghcr.io/seanmorley15/adven | 6 |
|
||||
| R-635 | P3 | FIXED-BY-LATER-WORK | The open remainder (app_oom fires once per container run, no escalation) was built in felhom-controller 0054d4b (v0.265.0, 'OOM storm alarm', R-636). controller/internal/notify/notifier.go:746 '\t\tn.emit("app_oom_storm", "error",' fires once per run when 20 or more kills land in 30 min (:754-764, oomStormKills=20, oomStormWindowMin=30; pinned by TestR636_*). The 79 % headroom and the method lesso | 6 |
|
||||
| R-645 | P3 | STILL-TRUE-NOT-SMALL | Still true for the operator CLI: controller/internal/settings/settings.go:1923-1929 ClearRestoreHold deletes ANY hold reason (including update-failed) with no pin restore, and the flag is in controller/cmd/controller/main.go:94/198. The row lists three candidate shapes with 'none chosen'. Picking one changes operator-path semantics and the capture logic, so it is not a one-hour fix. (The automatic | 6 |
|
||||
| R-675 | P3 | STILL-TRUE-SMALL | FIX: In missingFileLegsRefusal, when the second drive holds a whole copy of the app (the decision-26 whole-copy check the backup manager already exposes for the restore page), name that whole restore action instead of 'Fájlok visszaállítása'. Keep the existing branch otherwise. — Unchanged: controller/internal/web/handlers.go:1745 'return head + "A fájlok a második meghajtó másolatából állíthatók | 5 |
|
||||
| R-676 | P3 | STILL-TRUE-NOT-SMALL | A watch row, still true: controller/internal/stacks/unhealthy.go:116 skips only 'if st.Deploying // st.Updating // st.HoldReason != "" // st.updateHeld {', so a deploy's first start (after Deploying clears) is sampled by decision 28's crash-loop stop. The immich cause is fixed in the catalog (56c4888, 768M, per the row). Covering a slow first start would need a first-start grace decision. | 5 |
|
||||
| R-682 | P3 | STILL-TRUE-NOT-SMALL | felhom-controller 7690c27 (v0.296.0): no remove journal in source — grep -rni 'remove.*journal/removeJournal/remove_intent/interrupted remove' controller/*.go returns nothing; git log --grep R-682 empty. Needs a new boot-time journal mechanism. | 3 |
|
||||
| R-683 | P3 | UNCHECKED | Watch item about a power-cut drill outcome; behaviour only a live box shows; git log --grep R-683 empty in controller. | 1 |
|
||||
| R-698 | P3 | NOT-WORTH-IT | PICK close-as-accepted (option a), operator to confirm: A backup records the image name/digest, not the image; restoring a version the maker deleted from the registry fails at the pull. | 2 |
|
||||
| R-700 | P3 | FIXED-BY-LATER-WORK | felhom-controller 820e8ef (v0.276.0, R-697/R-700). controller/internal/stacks/migrate.go:789: 'm.logger.Printf("[INFO] [stacks] %s: data moved %s -> %s — app.yaml keeps its pin (%d service(s)) and records"' — persistDriveFlip (migrate.go:763) loads app.yaml and changes only HDD_PATH. Only a live proof on a two-drive box remains (row's own residue). | 3 |
|
||||
| R-704 | P3 | FIXED-BY-LATER-WORK | felhom-controller 7cba0bf (v0.278.0). controller/internal/api/router.go:918 'func (r *Router) dropLeftoverHold(name, why string) {' calling r.sett.ClearUpdateHold(name); pinned by TestR704_AFreshInstallDropsALeftoverHold. Residue: live proof of the install-time drop only. | 2 |
|
||||
| R-706 | P3 | FIXED-BY-LATER-WORK | felhom-controller 0c702f8 (v0.279.0). controller/internal/api/router.go:946 'if err := r.backupMgr.DeleteOffsiteRestoreCopy(name); err != nil {' inside removeVerificationCopy (R-706); pinned by TestR706_RemovalWithBackupsDeletesTheVerificationCopy. Residue: not seen live. | 2 |
|
||||
| R-717 | P3 | STILL-TRUE-NOT-SMALL | app-catalog templates/opengist/.felhom.yml:42 and templates/wishlist/.felhom.yml:42 carry only signup_block; no after_setup/after_install in either .felhom.yml (grep empty). Fix needs per-app DB writes (opengist sqlite with app stopped) plus live proof. | 3 |
|
||||
| R-723 | P3 | FIXED-BY-LATER-WORK | felhom.eu 80aeac71 (hub v0.126.0). hub/internal/monitor/staleness.go:174 '// R-723 (v0.126.0): a customer's NEW box is not a recovery.'; hub/internal/notify/dispatcher.go:540 '"suppressed", "first hour of a new box (R-723)", "operator"'. Residue: live proof at a real first install. | 2 |
|
||||
| R-724 | P3 | STILL-TRUE-NOT-SMALL | Text parts fixed in controller 6be6c53 (v0.283.0). Remaining LAN/gateway read still goes only through the samba container: controller/internal/stacks/guestnet.go:47 'out, err := dockerexec.Command("docker", append([]string{"exec", sambaContainer}, args...)...).Output()' — the comment (guestnet.go:18) accepts that reads fail while sharing is off. Needs another read path (design). | 3 |
|
||||
| R-728 | P3 | STILL-TRUE-SMALL | FIX: Add a per-customerID in-flight guard (sync.Map/mutex set) around the create handler from the duplicate check to the self-bind mint, so a second concurrent submit for the same ID gets the 'already exists' form; optionally disable the submit button on submit in the template. — No commit fixes R-728 (git log --grep in felhom.eu only the filing commit 11591f3a). hub/internal/web/configs.go:705 'e | 5 |
|
||||
| R-729 | P3 | STILL-TRUE-NOT-SMALL | No route clears the off-site target: controller/internal/web/server.go:744-783 lists /backup/offbox/{config,toggle,enable-all,offer-dismiss,run,reset,status,restore,place,reconstitute,verify-copy/delete,confirm-escrow,inject-password}; /reset (offbox_handlers.go:311) only resets an orphaned repo. New press needs handler + settings clear + template + HU/EN copy + escrow/hub-managed-target interplay | 4 |
|
||||
| R-733 | P3 | STILL-TRUE-NOT-SMALL | Harness/golden-evidence change plus a decision whether proofs run with swap off; no commit references R-733. Not a source-verifiable single fix. | 1 |
|
||||
| R-738 | P3 | NOT-WORTH-IT | PICK close-as-accepted (wger defect fixed; the generic gap is a design note): The guarded Update's health check sees only an app's front page, so an app that serves its front page while its data is broken passes as done. | 3 |
|
||||
| R-747 | P3 | STILL-TRUE-NOT-SMALL | Lockout shortened: app-catalog a4597cd; templates/mealie/docker-compose.yml:27 ' - SECURITY_USER_LOCKOUT_TIME=1'. Residue still open: hourly lock renewal by a stranger (needs decision 57 option d) and page copy; needs decision + live proof. | 2 |
|
||||
| R-755 | P3 | DUPLICATE | of R-762 — Still true: templates/wger/docker-compose.yml has no WGER_USE_GUNICORN (grep empty). R-762 (open, read) states 'Owner decides together with R-755 (same server question)' and its fix names 'the gunicorn switch of R-755'. | 2 |
|
||||
| R-756 | P3 | UNCHECKED | Depends on whether 9202's scratch drive is a registered drive — live box state; the row itself says not measured which. Not verifiable from source. | 1 |
|
||||
| R-757 | P3 | STILL-TRUE-NOT-SMALL | No commit references R-757. controller/internal/stacks/deploy.go:1339-1349: 'case "secret":' ... 'value, err := generateValue(field.Generate)' ... 'appCfg.Env[field.EnvVar] = value' for any missing field of a deployed app, with no exception for fields consumed only by after_install. Fix needs a design (a marker for given-at-install fields or an ask path). | 3 |
|
||||
| R-758 | P3 | STILL-TRUE-NOT-SMALL | Still true: templates/bookstack/.felhom.yml:18 ' mem_limit: "512M"' vs compose limits docker-compose.yml:42 '512M' + :79 '256M'; onboarding/EXISTING-APPS-GAPS.md:26 still lists all 8. No gate (only scripts/onboarding_gaps.py:171 reports it). Not small: raising 8 figures changes the capacity check (decision 22) — which apps fit a box — plus a gate with decoy and a publish. | 4 |
|
||||
| R-762 | P3 | STILL-TRUE-NOT-SMALL | templates/wger/docker-compose.yml sets no DJANGO_DEBUG and no static/media server (grep empty); templates/wger/.felhom.yml:18 'lifecycle: hidden' (catalog 55b8c8a). Needs a server design decision (nginx sidecar vs gunicorn+static) and bench+box proof. | 2 |
|
||||
| R-763 | P3 | STILL-TRUE-NOT-SMALL | templates/wger/docker-compose.yml sets neither ALLOW_REGISTRATION nor ALLOW_GUEST_USERS (grep empty); wger hidden (.felhom.yml:18 'lifecycle: hidden'). The env change is tiny but its proof needs a working wger on 9202 (blocked by R-762); best done in the same session as R-762. | 2 |
|
||||
| R-774 | P3 | STILL-TRUE-NOT-SMALL | templates/karakeep has no Sentry/phone-app sentence (grep -i sentry empty); mail-ON proof needs a hub-enabled live box (demo-hp) — live work. | 2 |
|
||||
| R-775 | P3 | STILL-TRUE-NOT-SMALL | Narrowed (Grimmory published behind the family gate). Residue: per-name 15-min lock is hard-coded upstream (no setting) and the reinstall-over-kept-books finding is uninvestigated — needs live investigation. | 2 |
|
||||
| R-776 | P3 | STILL-TRUE-NOT-SMALL | Only bookstack has it: templates/bookstack/docker-compose.yml:32 ' - APP_PROXIES=172.16.0.0/12'; grep for TRUSTED_PROXIES/CORE_TRUST_PROXY/IPEXTRACTION/N8N_PROXY_HOPS in kimai, zipline, vikunja, nextcloud, n8n compose returns nothing. Five apps, each needing a live 3.6 re-measure on 9202. | 3 |
|
||||
| R-778 | P3 | NOT-WORTH-IT | PICK close-as-accepted: If a box rolls back to a controller older than 0.286, the dashboard's login counter trusts the leftmost forwarded address and can be dodged until the box moves forward. | 2 |
|
||||
| R-782 | P3 | STILL-TRUE-NOT-SMALL | Source agrees: templates/glance/docker-compose.yml:27 seeds glance.yml with no auth: block (grep 'auth' in templates/glance empty); templates/homepage has no HOMEPAGE_ALLOWED_HOSTS (grep empty). Needs live measurement on 9202 and a decision whether a public glance dashboard is intended. | 3 |
|
||||
| R-783 | P3 | NOT-WORTH-IT | PICK close-as-accepted (re-open if upstream exposes the setting): Three wrong SparkyFitness sign-ins by anyone block every visitor's sign-in for about 10 seconds. | 2 |
|
||||
| R-785 | P3 | STILL-TRUE-NOT-SMALL | templates/sparkyfitness/docker-compose.yml:45 ' image: codewithcj/sparkyfitness_server:v0.17.3' and :89 'codewithcj/sparkyfitness:v0.17.3'. Major-version ladder walk (bench + box), gated on R-784. | 1 |
|
||||
| R-831 | P3 | NOT-WORTH-IT | PICK keep (rotation is the operator's call; do not close a leaked-secret row silently): The Hetzner storage API token was printed into one session transcript. | 1 |
|
||||
| R-836 | P3 | STILL-TRUE-NOT-SMALL | Live host boot-loader work needing operator-approved reboots and measurement (GRUB env block on ESP, sp5100_tco arming). | 1 |
|
||||
| R-839 | P3 | STILL-TRUE-NOT-SMALL | Gate behaves as described: controller/cmd/controller/main.go:2397 'return false, "drive " + hdd + " is not a live mountpoint"' with hdd = cfg.Env["HDD_PATH"] (main.go:2379). Which writer put a per-app path in HDD_PATH is undiagnosed — a diagnosis task, not a one-hour fix. | 4 |
|
||||
| R-853 | P3 | NOT-WORTH-IT | PICK close-as-accepted: After a boot the box's versions and crash facts reach the hub up to about 15 minutes late. | 4 |
|
||||
| R-862 | P3 | UNCHECKED | Waiting on the operator's by-hand bootstrap on Tester 2 through his tunnel; whether done is live-box state. No commit records it (felhom.eu log since 2026-10-04). | 1 |
|
||||
| R-870 | P3 | NOT-WORTH-IT | PICK keep (operator's call; close when Tester 1 is retired): Tester 1's two Cloudflare tokens (disposable test customer) were printed into one session transcript. | 1 |
|
||||
| R-879 | P3 | STILL-TRUE-NOT-SMALL | hub/internal/store/store.go:166 'retrieval_password TEXT NOT NULL,' and store.go:1586/1592 write retrieval_password and api_key as given; no seal on them (grep seal near these fields empty). Sealing/hashing three tables with migration is a security change, more than an hour. | 3 |
|
||||
| R-882 | P3 | UNCHECKED | Longhorn instance-manager state on DooPlex (Tier 2, forbidden to touch); live-only, owner operator. | 1 |
|
||||
| R-883 | P3 | UNCHECKED | homelab-manifests repo is not in this workspace (ls /mnt/5_hdd/felhom.eu/git shows only app-catalog-felhom.eu, drills, felhom-agent, felhom-controller, felhom.eu); live DooPlex check (kubectl) is out of scope. | 1 |
|
||||
| R-886 | P3 | UNCHECKED | DooPlex Alertmanager volume ownership; homelab-manifests not in this workspace and live check not permitted. | 1 |
|
||||
@@ -33,6 +33,35 @@ The full text of every row below: `git show ab2b3049:documentation/backlog/OPEN-
|
||||
| Row | What | Closed | Evidence |
|
||||
|---|---|---|---|
|
||||
| **R-885** | **The Python tests under `felhom.eu/scripts/` did not run in CI (P4).** New gate `script-tests` (`scripts/script_tests_gate.py`) walks `scripts/` and runs every `test_*.py` on every push (13 suites, ~20 s); a suite's verdict is its exit code; finding none fails; nested runs step aside. The hub-DB script tests use a Python-sqlite3 stand-in when the CI runner has no `sqlite3` CLI (said out loud). Decoys (5) declared in `test_gate_decoys.py`; red-proofs R885-a/b convict. | CLOSED 2026-10-05 — FIXED (gate) | `scripts/script_tests_gate.py`; `scripts/test_script_tests_gate.py`; `audits/burndown-2026-10-05/r885-red-proof.txt` |
|
||||
| **R-184** | **Nothing prevents the hub from vouching an agent version that was never released.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Fixed by felhom.eu b55fc17d "hub v0.102.0 — refuse to vouch a version that cannot be installed (R-273)" — exactly shape (b), validate at vouch time in the hub. felhom.eu/hub/internal/web/configs.go:1358 `res := s.gitea.PackageDownloadable(ctx, t.pkg, t.version, t.file)` and :1365 `s.logger.Printf("[WARN] artifact vouch REFUSED: %s package %s is NOT downloadable (R-287)", ...)`; tag leg at :1343 TagServesFile; unreachable registry also refuses. |
|
||||
| **R-207** | **`DRY_RUN=1` on `node-housekeeping.sh` is NOT non-mutating — it destroys the metric history it is supposed to let you inspect** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Fixed by homelab-manifests fc9fbb8 "node_housekeeping: guard DRY_RUN, correct the expired Docker rationale, pin container log rotation". /home/kisfenyo/git/homelab-manifests/homelab-ansible/roles/node_housekeeping/templates/node-housekeeping.sh.j2:137 `if [[ "${DRY_RUN}" == "1" ]]; then` inside write_metrics, :138 logs "file left untouched". |
|
||||
| **R-287** | **`felhom-agent` CI is red for a TRUE reason, and the diagnosis it was filed under is wrong in every particular.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Deleter established 2026-08-10 (R-267 newest-10 prune, recorded in the row itself); CI fixed by felhom-agent 53d047a "Two guards, one number: bound the published check to the retention it must live with" (R-291). felhom-agent@e06ed97 scripts/check-published-versions.py:101 `RETENTION_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "retention-policy.json")`, :213 `keep = retention_kept()`; scripts/retention-policy.json:37 `"generic_versions_kept": 10,`. Follow-up row R-291 is open (OPEN-ITEMS.md:439). |
|
||||
| **R-289** | **R-182's register row describes a defect the code no longer has — an OPEN row that is a false alarm.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | R-182 was closed by felhom.eu ef6ac6fe (2026-08-22, register compression): documentation/backlog/CLOSED-ITEMS.md:474 `/ **R-182** / ... / **CLOSED — SHIPPED** (controller v0.194.0 + hub v0.90.0/.1, 2026-08-03) /`. The residue (digest never seen delivering) was since observed: documentation/audits/DRILL-chaos-night-2026-09-17.md:181 `backup_run_failures` „1 of 12 apps failed to back up in this nightly run: nextcloud" listed as an alarm that fired and was true. |
|
||||
| **R-373** | **`SysDataGrowGB` is the intended lever for the system-data volume, it works, and nothing sets it.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Premise (20G/50G two-volume mismatch, 'nothing sets SysDataGrowGB') was retired by agent v0.120.0 one-data-volume work, commit cd6e267 'v0.120.0 — one data volume (R-165...)'. felhom-agent/internal/reconcile/bringup.go:191 '// SysDataGrowGB is a COMPATIBILITY INPUT since agent v0.120.0 (R-165). There is no longer a second' and :437 'growGB := spec.DataVolGrowGB + spec.SysDataGrowGB'; installer passes it (felhom-host-install.sh:3140 '-sysdata-grow "$SYSDATA_GROW"') and records the sizing at :2041-2058. Note: cd6e267 predates the row's filing date; the row quoted an older audit. |
|
||||
| **R-390** | **The golden-bake runbook omits `pveam update`, and the failure it produces names the wrong cause.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Commit 2344589a ('... runbook pveam note'); felhom.eu/documentation/runbooks/RUNBOOK-manual-build.md:154 '2. Run **`pveam update` first** — the `virgin` snapshot's template INDEX is stale too, and a stale index fails as a bogus'. |
|
||||
| **R-427** | **`closed_register_gate.py` checks ONE direction only: an open word in a CLOSED row. The mirror — a CLOSED verdict on a row still sitting in `OPEN-ITEMS.md` — is unchecked, and there are TWELVE.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Commit 71b8c8c6 (Backlog triage Part B: '... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS'); felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word (CLOSED, SHIPPED,'. Of the 12 named rows, R-385/387/341/378/405/88a/88b/123 are now only in CLOSED-ITEMS.md; R-190 and R-352 remain open (partly-closed, as the row predicted). |
|
||||
| **R-437** | **The register compression sweep is OWED, and it was deliberately NOT run inside the 2026-09-01 beta-line session — this row is the record of that choice, not a note.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu 71b8c8c6 'Backlog triage Part B: 125 finished rows + 20 id-less rows moved to CLOSED-ITEMS ... closed_register_gate RULE 3 refuses a finished row in OPEN-ITEMS (decoys, seen red) ... register 444 -> 325'. felhom.eu/scripts/closed_register_gate.py:10 'RULE 3 — (2026-10-03) no row in `OPEN-ITEMS.md` may carry a CLOSED-family word'. Gate run today: 'closed-register gate OK — no open work filed as closed, no id in both registers.' (456 closed / 336 open, 0 convicted). |
|
||||
| **R-464** | **[P3-LOW] MariaDB's entrypoint prints `MariaDB upgrade not required` on an UNSUPPORTED DOWNGRADE, so that line cannot be used as a soundness signal.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Lesson homed and harness uses the correct probe. app-catalog-felhom.eu b7ef0c4 'upgrade-test.py: record the engine's own view of its datadir'; app-catalog-felhom.eu/scripts/upgrade-test.py:278 '"mariadb-upgrade --check-if-upgrade-is-needed --user=root "'. felhom.eu d6837d98 (SPIKE R-459); felhom.eu/documentation/architecture/09-update-architecture.md:1647 '1. **Ask the engine, not the log.** MariaDB's entrypoint prints `MariaDB upgrade not required` on an' (cites R-464). |
|
||||
| **R-501** | **[P3-LOW] The documented "confirm your CI run" recipe reads only the LAST page of the jobs list, and that list is not in id order — so it can report a run as missing that exists and passed.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu a4993272 'CLAUDE.md: the CI-check recipe was wrong in two ways, both measured today'. felhom.eu/CLAUDE.md:176 'rows — a run can sit several pages earlier. **Scan every page** and match on `head_sha`; with a'; recipe at CLAUDE.md:166-168 loops every page. |
|
||||
| **R-602** | **[P3-LOW] The language a signed-in page uses is NOT the language a cookie asks for, and a live probe that forgets this reports a fixed defect as unfixed.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu e02bc038 'hub v0.119.0 — ... R-596/R-598 closed' added the finding; felhom.eu/documentation/architecture/10-localisation.md:809-812 'the `felhom_lang` cookie and got the **Hungarian** page for `en`. ... The cookie is the right instrument for the anonymous claim page and the **wrong**' and :509 '`langFor`'s order is fixed: `?lang=` → **the household's setting when a session exists**'. Only the optional pointer from the workspace live-validation rules is absent (grep ?lang=/felhom_lang in CLAUDE.md files and .claude/rules: none) — a 5-minute add if wanted. |
|
||||
| **R-617** | **[P3-LOW] The Gitea API token this project uses for pushes cannot create a repository through the documented endpoint, but CAN through `repos/migrate` — so "the token cannot do it" was nearly recorded as a fact when the truth was "one endpoint refuses it".** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu@462ab4a5 (2026-09-22) documentation/architecture/09-update-architecture.md:1969 "`POST /api/v1/repos/migrate` is the route that works (the project's Gitea tokens carry" - continues at :1970 "`write:repository` but not `write:user`, so `POST /user/repos` answers 403"; recipe at :1979. The one-line note the row asked for exists (in the architecture doc rather than operations/). The optional operator-scoped token is a separate wish, not the defect. |
|
||||
| **R-705** | **[P3-LOW] There is no way to run the night's chain now — only its pieces.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | The remaining half (manual whole-guest backup) EXISTS and predates the row: felhom-controller bbed5af (v0.47.0, 2026-06-12) 'backups page — whole-guest backup visibility + manual trigger'. Live source @7690c27: controller/internal/web/backup_handlers.go:324 `case r.URL.Path == "/api/guest-backup/trigger" && r.Method == http.MethodPost:`; :340 `if err := s.backupTrigger.TriggerNow(); err != nil {`; quiesce.go:427 "manual backup requested — quiescing now" (bypasses due-ness, all tiers); wired cmd/controller/main.go:2099 and the page button backups.html:229. Agent side: felhom-agent internal/localapi/server.go:514 `mux.HandleFunc("POST /backup", ...)`. The controller half was built v0.279.0 (night-chain, handler_debug.go:79). The row's 'no manual trigger' claim was not true at writing; it is not chained into night-chain, which the row did not require. |
|
||||
| **R-766** | **[P3-LOW] A new app's logo and screenshots reach the boxes only with the next HUB release — the website alone is not enough.** (P4) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Hub releases after the assets push (felhom.eu 40f07429, 2026-10-01): hub v0.131.0 (2026-10-04) .. v0.136.0 (d4be9f6f, 2026-10-05). The build copies website assets: felhom.eu scripts/build-hub.sh:98 `cp "${WEBSITE_ASSETS_DIR}"/*-logo.svg "${BUILD_DIR}/assets/" 2>/dev/null // true`, and the hub build workspace /mnt/5_hdd/felhom.eu/build/felhom-hub/workspace/assets/ holds radicale-logo.svg + 3 screenshots (also karakeep, dawarich) dated Oct 5 14:28; hub/Dockerfile:27 `COPY assets/ /usr/share/felhom/assets-seed/`. Not checked: what a live box shows (no machine access). Checked 2026-10-05: the live hub image (v0.136.0) `/usr/share/felhom/assets-seed/` holds radicale-, karakeep- and dawarich-logo.svg + screenshots. |
|
||||
| **R-50b** | **[P2] A root-owned privileged host artifact is delivered unversioned from `main` — "which wrapper is on this host?" is unanswerable.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | Claim 'fetched via fetch_raw from raw/branch/main — no tag, no pin' no longer true: bee68484 (installer v1.23.0, R-110/R-183) pinned fetch_raw to the vouched agent tag — felhom.eu scripts/felhom-host-install.sh:533 '"$GITEA_BASE/$GITEA_OWNER/$AGENT_REPO/raw/tag/v$ART_AGENT_VER/$path" \' (leg b). Leg (c)-like signed delivery: felhom-agent c9fa2e7 (R-840 config bundle) and configs/test_felhom_config_bundle.py:264 covers /usr/local/sbin/felhom-pbs-apply. Residual worth one line if kept: the 0440 sudoers drift visibility note. |
|
||||
| **R-121** | **A BOX's installed agent can sit releases behind the vouched one and nothing notices — the R-120 gate does not cover it.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | 3d7a2761 (hub v0.135.0, R-530/R-604 'boxes left behind listed and alarmed'): felhom.eu hub/internal/osupdates/service.go:79 'EventAgentBehind = "agent_behind" // warning, operator' with :180 'AgentBehindAfter: a box runs an agent older than the vouched one this long → an operator alarm' (7 d window, the staleness window the row asked for). |
|
||||
| **R-200** | **The DR password-injection seam has a handler, a route and tests — and no form.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | The remaining half (customer-facing recovery-code form: yell → R form → preview) shipped as the recovery screen: felhom-controller 636c51e 'R-193: the recovery screen — unlocking, and only unlocking (v0.200.0)'; controller/internal/web/templates/recovery.html:82 '<form id="unlock-form" method="POST" action="/recovery/unlock" autocomplete="off">', routed at internal/web/server.go:602. Plumbing half was 1b1366b (v0.196.0). The 64-hex inject-password route (server.go:783) stays a deliberate DR fallback with no form. |
|
||||
| **R-450** | **[P2-MEDIUM] UPDATE ARC SLICE 6 — a version sequence: automatic WITHIN a major, never ACROSS one, and an engine change gets its OWN edge.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | The row's only remainder was 'the other ten PostgreSQL apps need two-venue proof'. R-463 CLOSED 2026-09-30 by felhom.eu 25cb3eb9 ('The last six PostgreSQL apps decided'): 8 of 11 moved by the box's own conversion, 3 (zipline, adventurelog, immich) stay by decision 42; CLOSED-ITEMS.md:223. Source proof: app-catalog templates/docmost/docker-compose.yml:62 'image: postgres:18-alpine' (also rallly:67, outline:64, paperless-ngx:101). The per-app engine gate stays as the permanent rule (catalog CLAUDE.md:115-122). |
|
||||
| **R-489** | **[P3-LOW] `POST /api/stacks/{name}/remove` reports `volumes_removed: null` over named volumes it DID remove.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | The residual (a unit-restore-recreated volume has no compose label, so the remove answered []) was fixed in felhom-controller 206b035 (v0.268.0, R-658). controller/internal/stacks/delete.go:1089-1090: '// appVolumeSet is every volume the removal accounts for: the ones carrying the project label AND the // ones the app's definition declares that Docker holds by name (R-658, v0.268.0).' delete.go:703 'resp.VolumesRemoved = removedVolumes(volsBefore, m.appVolumeSet(name, stackDir))'. |
|
||||
| **R-573** | **[P3-LOW] The agent-channel and endpoint-drift banners reach the dashboard as finished Hungarian, so they stay Hungarian on an English page.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom-controller 7c4a33b (v0.258.0, 'the last four Hungarian things an English household met ... R-573 the two channel banners'). controller/internal/web/alerts.go:113 'func (am *AlertManager) SetAgentChannelAlert(down bool, msgKey, msg string) {' and :136 'func (am *AlertManager) SetEndpointDriftAlert(drift bool, msgKey, msg string) {'. Both set MessageKey with msg only as a fail-open fallback. |
|
||||
| **R-622** | **[P2-MEDIUM] `adventurelog v0.13.0` migrates the customer's database and then does not serve — the edge must NOT be promoted, and it is the first real-catalog candidate this project has measured as unsafe.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | adventurelog v0.13.0 was diagnosed, fixed and promoted with a two-venue test record in app-catalog-felhom.eu 06ea7da (2026-09-27, 'adventurelog: v0.12.1 -> v0.13.0 with its health and world-data fixes in the same commit (R-655, 09 decision 41)'; bench healthy in 217 s, box 9202 through the guarded Update in 204 s). templates/adventurelog/docker-compose.yml:13 ' image: ghcr.io/seanmorley15/adventurelog-backend:v0.13.0'. The proven step is recorded at templates/adventurelog/.felhom.yml:132 (update_ladder entry from v0.12.1). |
|
||||
| **R-635** | **[P1-HIGH] `romm 5.3.0` does not fit the memory the template gives it, and the guarded Update called that a success — the app has been OOM-crash-looping on demo-hp for six hours at ~500% CPU.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | The open remainder (app_oom fires once per container run, no escalation) was built in felhom-controller 0054d4b (v0.265.0, 'OOM storm alarm', R-636). controller/internal/notify/notifier.go:746 '\t\tn.emit("app_oom_storm", "error",' fires once per run when 20 or more kills land in 30 min (:754-764, oomStormKills=20, oomStormWindowMin=30; pinned by TestR636_*). The 79 % headroom and the method lesson are carried by R-462 (per the row). |
|
||||
| **R-235** | **The appliance console keeps telling an already-paired box to go and pair itself.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu c033b3b6 'ISO 1.28.0 source: the console stops showing the pairing code once bound (R-535)'. scripts/iso/felhom-bootstrap.sh:538: `print_bound_banner # R-535: replace the pairing code on the console with the truth`. Same defect already CLOSED twice in CLOSED-ITEMS.md as R-535 (line 232) and R-214 (line 200, 'proven on a fresh install'). |
|
||||
| **R-282** | **One secret, three different Hungarian names, and the email sends the customer to a page their box is not showing.** (P3) | CLOSED 2026-10-05 — FIXED BY LATER WORK (burn-down Part A) | felhom.eu 4d6ec7c 'hub v0.104.0: ... the hub half of the naming (R-295)' + controller v0.211.0 (R-295 CLOSED, CLOSED-ITEMS.md:207) + R-323 hub v0.105.0. hub/internal/notify/templates.go:204: '// R-295, HUB HALF (2026-08-13). ONE NAME PER SECRET, and it is „Beállító kód".'; hub/internal/claim/engine.go:51 `EmailReenroll EmailKind = "reenroll"` (mail names the setup page a rebuilt box shows). |
|
||||
| **R-376** | **The placement decision that cost four mis-filed defect reports was never recorded as a decision anywhere, and the architecture folder's own marker convention lives in one document of eight.** (P4) | CLOSED 2026-10-05 — DONE | The legend reached the three documents written after the 2026-08-22 pass (`08`, `09`, `11`); all eleven numbered architecture documents now carry it (`grep -c "not yet classified"` = 1 each; `07` is the original home). The hot/bulk decision was given its home on 2026-08-22 (`01` [DESIGN] + CONTEXT). Marking every statement stays a practice of each session that touches a document, not a defect. |
|
||||
| **R-817** | **`09` decision 56 and R-745 disagree about what the controller self-update rolls back to.** (P4) | CLOSED 2026-10-05 — CLARIFIED (no contradiction) | `felhom-agent internal/localapi/controllerswap.go:236-240` records the image running when a swap starts as `Previous`; `:289` writes it back on failure. After a good swap that is „the one before" the running image — decision 56 and R-745 name the same image. A dated clarification sits under decision 56 in `09`; the ruling text is unchanged. |
|
||||
| **R-818** | **Two changelogs cite register ids for other findings.** (P4) | CLOSED 2026-10-05 — CORRECTED | Dated correction notes under hub v0.109.0 (`hub/CHANGELOG.md`) and controller v0.224.0 + v0.225.0 (`felhom-controller/CHANGELOG.md`): those two findings never had register rows of their own — the triage's „the real ids are in CLOSED-ITEMS" was itself wrong (no closed row names hub v0.109.0 or controller v0.224.0/v0.225.0). Nothing renumbered. |
|
||||
| **R-755** | **[P3-LOW] wger runs Django's DEVELOPMENT server in production: `manage.py runserver`, because the template does not set `WGER_USE_GUNICORN=True`.** (P3) | CLOSED 2026-10-05 — DUPLICATE of R-762 (its unique fact moved there) | Still true: templates/wger/docker-compose.yml has no WGER_USE_GUNICORN (grep empty). R-762 (open, read) states 'Owner decides together with R-755 (same server question)' and its fix names 'the gunicorn switch of R-755'. |
|
||||
| **R-446** | **[P2-MEDIUM] „Naprakész" can be FALSE, and the badge that says it cannot tell.** (P3) | CLOSED 2026-10-05 — DUPLICATE of R-440 (its unique fact moved there) | felhom-controller/controller/internal/stacks/updateorder.go:96: `if len(s.CatalogDigests) == 0 // s.CatalogTestedAt.IsZero() { return false }` — blind only for apps with no ladder entry, i.e. the same 15 templates R-440 lists (app-catalog has no update_ladder for them). Both rows close by the same act: each app's first proven ladder step (R-462). |
|
||||
|
||||
---
|
||||
|
||||
|
||||
@@ -114,16 +114,14 @@ match what the reader sees is how an instrument stops being believed (R-421). It
|
||||
stopping line that lies.
|
||||
|
||||
|
||||
## Install & onboarding — 17 rows (P3 11, P4 6)
|
||||
## Install & onboarding — 15 rows (P3 9, P4 6)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
| **R-130** | Install & onboarding | P3 | **A "hard min" that only warns.** A fresh box's `local-lvm` was ~75 GiB against `HARD_MIN_LVM_GIB=120` (`scripts/felhom-host-install.sh`); the installer logged `[WARN] local-lvm free ~75 GiB < hard min 120 GiB` and went on to a **fully successful** install | READY (S) | — | Either the minimum is not hard (rename it and state the real floor) or it is wrong (and 120 GiB is not what a working appliance needs). Leaving it is the R-29 shape: a check that reads as coverage while providing none. Evidence: same audit §8 | CC |
|
||||
| **R-179** | Install & onboarding | P3 | **`--uninstall` leaves the NAS network-storage systemd units behind, with the automount in `failed` state and the parent bind still mounted.** The teardown's residue-diff provenance (`day0-install.md` Part E: *"a full-filesystem diff against the pre-install baseline showed zero Felhom-named leftovers"*) is from **v1.9.1**, which predates the NAS network-storage feature. A box that has ever had a network share configured keeps `/etc/systemd/system/mnt-felhom\x2ddrives-<share>.mount` and `.automount` after a full uninstall | **READY (S) — NEW 2026-08-03** | — | **Observed on demo-hp 2026-08-03** after `--uninstall --vmid 9201`: `mnt-felhom\x2ddrives-Felhom\x2dShare.automount` **loaded failed failed**, its `.mount` `loaded inactive dead`, and `mnt-felhom\x2ddrives.mount` still `active mounted` — the uninstall's own output had warned `/mnt/felhom-drives/Felhom-Share is busy — NOT forcing` and `/mnt/felhom-drives root bind left mounted`, which is correct behaviour (it never forces an unmount) but is not teardown. Cleared by hand before the reinstall: stop both units, remove both unit files, `daemon-reload`, unmount the autofs then the parent. **NEGATIVE CONTROL, same day:** demo-felhom's uninstall left **nothing** (`ls /etc/systemd/system | grep -i felhom` → only the unrelated `felhom-bootstrap.service`; no felhom mounts) — because that box had no network share configured. **So the residue is conditional on the feature having been used, which is exactly why a diff taken on a box that never used it reported clean.** `felhom-bootstrap.service` is NOT residue — it is the ISO first-boot unit, `disabled`+`inactive`, exactly-once and already fired | CC |
|
||||
| **R-180** | Install & onboarding | P3 | **`--archive-storage` is accepted without checking the agent's token will ever be granted on it, and the failure lands at step 8/8 — after the root@pam password has already been rotated.** `felhom-host-install.sh` validates the archive storage EXISTS (`pvesm status --storage`, `:1583`) and that the golden volid RESOLVES on it (`:1661`), both in pre-flight. It never checks that storage against the ACL set it is about to grant, which is the fixed default `local local-lvm felhom-pbs` (`--acl-storages`, which `runbooks/day0-install.md` tells the operator **not** to pass). A storage outside that set therefore passes every pre-flight gate and dies at the last step | **READY (S) — NEW 2026-08-03** | — | **Hit live on demo-hp 2026-08-03** during R-178 Phase A, self-inflicted and therefore a clean demonstration: the golden was staged on `felhom-backup` (the enrolled NVMe, where the box's vzdumps live) and `--archive-storage felhom-backup` passed. Pre-flight passed; steps 1–7 ran; step 8 returned `reconcile: bring-up restore: proxmox: POST /nodes/felhom-host/lxc -> HTTP 403: permission denied at /storage/felhom-backup (missing privilege Datastore.AllocateSpace)`. **The cost is the ORDER, not the error** — by the time it fires, step 2 has minted the PVE token, step 4b has **rotated root@pam and vaulted it** (so the old console password is already dead), and step 5 has installed the agent. Recovery was `--resume` after moving the golden to `local`, which worked cleanly. **This is statically checkable in pre-flight**: `ARCHIVE_STORAGE ∈ PVE_STORAGES` is a one-line assertion over two variables both known at `:1583`. Same class as R-29 — the checkable thing that nothing checks | CC |
|
||||
| **R-235** | Install & onboarding | P3 | **The appliance console keeps telling an already-paired box to go and pair itself.** Measured 2026-08-06 on VM 323: **25 minutes after** the operator bind, with the guest provisioned, the controller reporting 0.203.0 and the agent ONLINE, the physical console still displayed „Felhom — a doboz készen áll, és a **párosításra vár**" together with the now-spent pairing code `US3-6GP` — and, in the same panel, the promise **„Ez a képernyő magától frissül — nincs teendő a doboznál"**. It does not refresh. A customer looking at their screen is told the setup has not happened, and is told the screen would have updated if it had. Cosmetic in mechanism, not in effect: it invites the customer to re-pair a working box, or to call for help about a box that is already fine. Same family as R-234 — a surface asserting a state that stopped being true. | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-250** | Install & onboarding | P3 | **A customer create can fail fail-closed because the host-key scan ladder is shorter than the DNS/AAAA settle time.** Found 2026-08-07 creating the fifth walk's venue. `POST /configs/new` with off-site enabled provisions the Storage Box sub-account and then scans its SSH host key to pin it — **fail-closed by design** (`offsite.go:111-121`, *"don't serve a descriptor the controller can't verify"*), with `defaultScanBackoff` = 2+4+8+16+30 ≈ **60 s**, sized by its own comment to *"the observed DNS propagation lag"*. **Measured, both halves:** the first create exhausted the ladder — five `no such host`, then `dial tcp [2a01:4f8:bacc:2:200::d30]:23: connect: network is unreachable` (the name had just begun resolving, **AAAA-first, into a pod with no IPv6 route**) — and the hub logged `[ERROR] offsite provision for walk5`. An **identical second POST succeeded** ~70 s later on its own final rung (`shared already provisioned … (subaccount 285351)`). **Total settle ≈ 100 s against a 60 s budget.** The endpoint was never the problem: verified afterwards from both the node and the hub pod, by name, `SSH-2.0-OpenSSH_9.6p1`. **Two distinct things are wrong and should not be merged:** (1) the budget is sized against DNS *existence*, but what actually bit is the **AAAA-before-A** window, a different and longer phenomenon — lengthen the ladder and/or prefer the A record for the scan dial; (2) **the operator is told nothing actionable** — the create fails with a generic error, the remedy is "press it again", and nothing says so. The retry is genuinely safe (idempotent on the label; confirmed afterwards — `one_time_secrets` 1, one sub-account, no double-provision) **but that safety is invisible to the person deciding whether pressing again will double-charge them.** **Severity LOW-MEDIUM:** self-clearing, no data at risk — but it is the first thing a new customer's provisioning does and it fails looking like an outage. | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-282** | Install & onboarding | P3 | **One secret, three different Hungarian names, and the email sends the customer to a page their box is not showing.** Sending it from the hub is „**Visszaállító** kód küldése"; the email that arrives is subject „Jelszó-**visszaállítási** kód", body „**Visszaállító** kód: …", and it instructs *„Add meg a vezérlőpult »**Elfelejtett jelszó**« oldalán"*; the page the box actually serves is „A szerver **beállítása**" asking for a „**Beállító** kód". **A rebuilt box shows a SETUP page and the hub can only send a RESET mail** (because hub-side the customer is still `claimed_at 2026-07-21`), so the instruction names a route that does not exist on screen. **It does work if you ignore the instructions** — the reset code was accepted on the setup page (302 + session), so this is naming, not function. **It cost this session real time and one wasted code:** the operator supplied a 3-word Hungarian code believing it was the recovery code, because the hub calls the claim code „Visszaállító kód" and the ESCROW code is also „Visszaállító kód" — the only reliable discriminator is length (claim = 3 Hungarian words; recovery = **10** EFF-list words, and the recovery screen does say „(tíz szó)") | **VERIFY** (2026-10-03 triage: Naming fixed on box (R-295), hub and mail page reference (R-295 hub, reenroll mail) and R-323 — OPEN-ITEMS.md R-327 row (line 578)) — **READY (S) — NEW 2026-08-09** | — | Pick one name per secret and use it on all three surfaces; make the mail's page reference match what a rebuilt box actually shows | CC |
|
||||
| **R-283** | Install & onboarding | P3 | **After a rebuild the hub says "Claimed 18d ago" while the box serves its first-run setup page.** `customer_claims` for demo-hp still read `claimed_at 2026-07-21 16:29:25`, `generation 2`, `issued_at 2026-08-03` while the freshly provisioned guest — whose `settings.json` is new — correctly showed „A szerver beállítása". The two sides never reconcile: the hub's claim state survives a guest rebuild and the box's does not. Consequences: the operator's screen says the box is claimed when it is not, a resend produces a RESET code instead of a SETUP code (→ **R-282**), and any previously issued code fails with *„Hibás vagy lejárt kód"* — a message that is technically true and tells the customer nothing about the real cause, namely their own reinstall. Mirror image of **R-214/R-235** (an already-paired box still told to pair itself) | **READY (S) — NEW 2026-08-09** | — | Let a report from a box carrying no claim state clear the hub's, or show both sides on the operator page | CC |
|
||||
| **R-306** | Install & onboarding | P3 | **`--preflight-only` says "no state written" and writes state — with an answer that can be wrong.** `_state_put` short-circuits on `DRY_RUN` only (`felhom-host-install.sh:418`), so a preflight-only run creates `/var/lib/felhom-install/state.json`. Observed live: after a run whose banner read `PRE-FLIGHT PASS (mode=byo) — no state written, no install step executed`, the file existed containing `{"completed": [], "dnsmasq_preexisting": "yes"}`. Both the banner and the flag's own comment at line 226 assert the opposite. **The harm is not the file, it is the value**: the runbook recommends preflight-only first, then the same command without the flag, so on a box carrying a Felhom leftover the wrong ownership answer is baked in before the real install begins | **READY (S) — NEW 2026-08-12, RANK 3** | R-300, R-305 | Either make `_state_put` a no-op under `PREFLIGHT_ONLY` (and record ownership at install instead), or correct both claims. A comment asserting an invariant needs a test pinning it | CC |
|
||||
| **R-317** | Install & onboarding | P3 | **The agent decides whether to install dnsmasq by stat-ing a file the OTHER package owns.** `EnsureDnsmasq` (`felhom-agent/internal/lanresolver/lanresolver.go:105`) does `os.Stat("/usr/sbin/dnsmasq")` and skips the apt install when it exists — but that path is shipped by **`dnsmasq-base`**, while the systemd unit comes from **`dnsmasq`** (confirmed on the box: `dpkg -S /usr/sbin/dnsmasq` → `dnsmasq-base`; `dpkg -S /usr/lib/systemd/system/dnsmasq.service` → `dnsmasq`). So on any host carrying `dnsmasq-base` without `dnsmasq`, the agent skips the install and then runs `systemctl enable --now dnsmasq` against a unit that is not there: the resolver never comes up and the failure is a retried WARN in the journal rather than anything a customer or the install sees. **Pre-existing, NOT introduced by R-316** — but R-316 makes the shape reachable, because a host whose `dnsmasq-base` pre-dated Felhom now keeps it while `dnsmasq` is removed. R-316's uninstall says so explicitly instead of leaving it to be found from a silent resolver. **Ranked 2 (costs time), not 1:** the box installs fine, only LAN name resolution is missing | **READY (S) — NEW 2026-08-13** | R-316 | Probe what is actually needed — the unit or the `dnsmasq` package — rather than a path a sibling package owns. One-line change in the agent; deliberately NOT made here to keep this session to one repo | CC |
|
||||
@@ -136,11 +134,10 @@ stopping line that lies.
|
||||
| **R-725** | Install & onboarding | P4 | **[P3-LOW] Small copy slips on the first-hour path.** MEASURED 2026-09-29: the self-bind PAGE says the passphrase was received „a beállításkor" while the mail and console say „a Felhom üzemeltetőjétől" (R-497 unified the mail and console, not the page); the recovery-code wizard addresses the household formally („Írja fel", „adja meg") while every other screen says „te"; the console's linked banner ends „a doboz össze van kötve. V" (a stray glyph); a gated app answers a phone app's API call with English JSON „this app is waiting for its first setup" (the browser gets the Hungarian gate page). **FIXED 2026-09-30:** the recovery wizard speaks „te" (controller v0.283.0; formal ceiling 18 → 17); the bind page says „a Felhom üzemeltetőjétől kaptál" (hub v0.126.0). **NARROWED — remaining:** the console's stray „V" (the installer/agent's banner, not these repos' text); the gate's English JSON to a phone app (the app shows its own error; left, deliberately); and the expired bind page still says „kérj újat az ügyfélszolgálattól" ABOVE the new „Új linket kérek" button (hub copy, next hub release). | **NARROWED — three small copy items; owner: CC** **Re-ranked 2026-10-03: P3→P4: three small copy slips left; nothing blocks the household.** | — | — | CC |
|
||||
| **R-881** | Install & onboarding | P4 | **The installer's uninstall does not remove `/usr/local/sbin/felhom-priv-apply`** (added to the bundle by agent v0.146.1, R-861), and its comment still says the guest hook is agent-installed at runtime (`scripts/felhom-host-install.sh` ~line 1679) — since v0.146.1 the hook is a bundle file. Found 2026-10-05 while checking the installer against the new bundle. | READY | — | Add the file to the uninstall list and fix the comment at the next installer tag | CC |
|
||||
|
||||
## Apps & catalog — 36 rows (P3 15, P4 21)
|
||||
## Apps & catalog — 33 rows (P3 13, P4 20)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
| **R-489** | Apps & catalog | P3 | **[P3-LOW] `POST /api/stacks/{name}/remove` reports `volumes_removed: null` over named volumes it DID remove.** MEASURED 2026-09-13 on demo-hp five times (gokapi, actualbudget, adventurelog ×2, glance): `docker compose down --volumes` removed the app's named volumes (`docker volume ls` count 2 → 0) and the response carried `"volumes_removed":null`. The customer's confirmation dialog therefore cannot say what it deleted. Split out of R-474 (closed in v0.240.0 for the backups half). **Fix shape:** list the volumes before `down --volumes`, diff after, and report the difference (`[]` when none, never `null`). **PARTLY SHIPPED in v0.242.0 (`d698ce3`), measured live on 9202 the same night:** the difference is computed and a fresh compose-created volume IS reported (`["opengist_opengist_data"]`), but the listing filters on the compose project LABEL and a volume recreated by a unit restore (`docker volume create <name>`, `restore.go:154`) carries no labels — compose still removes it and the response says `[]` (`audits/v0242-2026-09-14/19-R489-cause.txt`). **Remaining fix:** list by the `<project>_` name prefix as well (union), or label the recreated volume as compose would. | **READY — rank P3-LOW; owner: CC (residual)** | — | — | CC |
|
||||
| **R-498** | Apps & catalog | P3 | **[P3-LOW] The „Első lépések" on 52 of 53 app pages tell a customer to open a literal `wiki.DOMAIN` — the placeholder is never filled in.** MEASURED 2026-09-14 on a fresh 0.242.0 box (drill step 5): BookStack's app page renders „Nyisd meg a **wiki.DOMAIN** címet a böngészőben", PrivateBin's „Nyisd meg a **paste.DOMAIN** címet". `grep -rl '\.DOMAIN c[íi]m' app-catalog-felhom.eu/templates/*/.felhom.yml` → **52 of 53** templates carry it in `first_steps`; the controller renders the string as text (`internal/stacks/metadata.go`). A stranger reading their first instruction meets a word that is not an address. **Fix shape:** substitute the stack's real `SUBDOMAIN.DOMAIN` at render time (one place in the controller), with a render test per template that fails on a literal `DOMAIN`. | **READY — rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-562** | Apps & catalog | P3 | **[P3-LOW] Dates and sizes are not formatted for any locale — and the Hungarian pages disagree with themselves.** FOUND 2026-09-17 by the i18n inventory §2.8: the two template date layouts differ (`2006. 01. 02. 15:04` Hungarian vs `2006-01-02 15:04` ISO); 10 layout literals in `internal/web` Go and 25 elsewhere pick formats ad hoc; sizes print a decimal POINT (`%.1f GB`, 4 helpers) where Hungarian uses a comma; `timeAgo`/`nextRunLabel`/`pruneLabel` produce Hungarian words outside the three converted pages. Not changed by v0.247.0 (Hungarian bytes are frozen by the parity rule). **Fix shape:** one date and one size formatter per language in `internal/i18n`, the Hungarian output deliberately changed in ONE reviewed release with the parity fixtures re-captured for that release only and the change named in its CHANGELOG. Needs an operator word on the Hungarian format (comma, date style). | **READY - rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-575** | Apps & catalog | P3 | **[P3-LOW] The soft memory-overcommit warning is returned as a Hungarian STRING, so it renders Hungarian on an English page.** FOUND 2026-09-18 by localisation slice 2 release B (R-557, controller v0.253.0): `memoryVerdict` now returns its REFUSAL as an error carrying a key (`util.MsgErrorf(ErrNotEnoughMemory, …)`), but its WARNING is a plain string with no error to carry one, and no language is known where it is built — so it goes through `msgHU` and is always Hungarian. The copy is in the bundle (a translator sees it; `i18n_go_parity.py` pins it), only the render is fixed to `hu`. The named helper and this consequence are stated in `internal/stacks/deploy_errors.go`. **Fix shape:** the caller carries the key the way `Alert` and `UpdateRefusal` now do — `memoryVerdict` returns `(refusal error, warningKey string, warningArgs []any)` and the deploy answer renders it — not a helper guessing a language it cannot know. | **READY - rank P3-LOW; owner: CC** | — | — | CC |
|
||||
@@ -150,10 +147,9 @@ stopping line that lies.
|
||||
| **R-676** | Apps & catalog | P3 | **[P3-LOW] Watch: immich's first start restarted 12 times — decision 28's crash-loop stop (6 in 10 min) would stop it.** From the 2026-09-17 chaos night (DB connection dropped during the first-start geocoding import on a 6 GB guest; it did not recover that night). No healthy app in any drill evidence restarts on a first start (1831 samples, 40 live containers), so the threshold stands; this row exists so the first immich install under v0.269.x is watched. `audits/night-2026-09-24/A3/40-first-start-restarts.txt` **2026-09-25 night (read from source, v0.271.0): a DEPLOY's first start is NOT covered by decision 28's suppression** — `Deploying` clears when `compose up -d` returns (`deploy.go` "Clear deploying flag"), and `ObserveUnhealthy` then samples the app; an automatic update's step, verify and undo ARE covered (`Updating`, pinned by `TestD28_NoCrashLoopStopDuringAnAutomaticStep`). So a first start that restarts ≥ 6 times in 10 min is stopped — which R-676 already accepts for a broken first start; a healthy slow first start would be stopped too. **-- 2026-09-30: the first-start restarts are explained.** immich's first-start geodata import OOM-kills its database at 512M on a guest with no swap (R-732, measured: 61–104 kills); the 2026-09-17 chaos-night case (DB connection dropped during the import on a 6 GB guest) fits it. Fixed in the catalog (`56c4888`, 768M). The watch itself (decision 28 on a DEPLOY's first start) is unchanged. | **OPEN — P3; owner: CC (watch)** | — | — | CC |
|
||||
| **R-682** | Apps & catalog | P3 | **[P3-LOW] A Remove interrupted by a controller kill leaves the app half-removed: containers gone, the app still listed as installed (and held).** MEASURED 2026-09-24 on 9202 (chaos round 9): the kill 2 s after the Remove press answered the household `502 Bad Gateway`; after the restart `chaoscrash` read deployed, stopped, `unhealthy_stop`, with NO container left. Pressing Remove again completed it cleanly (200, only the catalog template left). Recoverable by the household's own second press; nothing tells them to press it. **Fix direction:** the remove journals its intent and finishes (or says it was interrupted) at boot, as the update does. `audits/night-2026-09-24/E/round-09*.json`, `E/round-09b-remove-again.txt` | **READY — P3; owner: CC (controller)** | — | — | CC |
|
||||
| **R-704** | Apps & catalog | P3 | **[P3-LOW] The box's crash-loop stop (decision 28) outlives the app: after remove and reinstall, the new install is still held.** Measured 2026-09-28 on 9202: calcom crash-looped at 08:22 and 08:28 (`unhealthy_stop`, `crash_loop`, trip 2, recorded 08:28:59Z); it was then REMOVED through the product twice and installed fresh twice (09:14:51Z the last). At 09:45 the new, healthy install's Update was refused `409 held` with the crash-loop sentence („…újra és újra összeomlott…"), and `GET /api/stacks/calcom` carried the old `hold_reason` while `state=running`. Start lifted it (`the unhealthy stop is LIFTED by Start`). So a household that removes a crash-looping app and installs it again (the obvious fix) finds its updates refused for a crash of a previous install. Not measured: whether the nightly update leg also skips it; whether other holds (restore hold) behave the same. **Fix direction:** the remove clears the app's box-set holds, as `DeleteAppBackupPrefs` clears its backup preferences (R-474). Evidence: `audits/pg-calcom-claper-2026-09-28/box/calcom/hold.txt`, `…/box/calcom/move.txt`. **-- 2026-09-28 later: SECOND and worse instance, then FIXED in controller v0.278.0.** demo-hp's fresh nextcloud (installed 10:13) carried an UPDATE hold from a nextcloud of 2026-09-13 (set before v0.242.0 made removals clear update holds; nothing ever swept it). At the manual off-site run (15:17) the backup leg logged `Skipping volume dump for nextcloud — the app is HELD stopped`, captured no unit, and pushed a snapshot that `carried NO database dump and NO volume tar` — a freshly installed app silently NOT backed up. **Fix:** a removal also clears the crash-loop stop (`settings.ClearUpdateHold`), and a new install (plain or "use my kept data") drops a leftover update/crash-loop hold of an app that is not installed (`Router.dropLeftoverHold`); restore holds (R-379) untouched. Red-proofed RP4–RP6 (`audits/kept-offsite-2026-09-28/redproofs/`). Floor 0.278.0. **STILL OPEN: live proof of the install-time drop** (a box with a leftover hold on an uninstalled app). | **WATCHING — P2; owner: CC (install-time drop, live)** | — | — | CC |
|
||||
| **R-755** | Apps & catalog | P3 | **[P3-LOW] wger runs Django's DEVELOPMENT server in production: `manage.py runserver`, because the template does not set `WGER_USE_GUNICORN=True`.** MEASURED 2026-10-01 on 9202 (`ps` in the wger container: `python3 manage.py runserver 0.0.0.0:8000`); wger 2.7's `extras/docker/production/entrypoint.sh:81-87` runs gunicorn only with that switch. Django's own documentation says runserver is not for production (one process, not hardened). Not changed this session (a different change from R-752's; needs its own bench + box proof, memory watch included). **-- 2026-10-01 (checklist pilot):** re-measured on 9202 at the live template: still `manage.py runserver` (`audits/new-app-checklist-2026-10-01/C/C2-static-reads.txt`). The same server question now carries R-762 — runserver with DEBUG off serves no `/static` and no `/media`, so the fix for both is one decision (upstream's nginx front, or gunicorn + a static server). | **OPEN — rank P3-LOW; owner: CC (catalog)** | — | — | CC |
|
||||
| **R-757** | Apps & catalog | P3 | **[P3-LOW] A template that gains a generated `secret` field makes the box INVENT that value for apps already installed — for calibre-web a login name the app never got.** MEASURED 2026-10-01 on demo-hp: 9 minutes after catalog `e9f50b5` (decision 61) synced, the controller logged `InjectMissingFields … injected missing fields: ADMIN_USER` (deploy.go:1337) and the app page's reveal returned a 10-character name that was NOT calibre-web's login (the app had kept its own name; `after_install` runs only after a fresh install). The page did not list the field, but the reveal answers it, and the password field's text now says "the user name above". demo-hp was fixed by renaming the app's user to the box's recorded name (credentials file updated). Any other installed calibre-web gets the same made-up name at its next sync while its login stays `admin` (no other box has one today: N100 none, Tester-2 not registered). **Needs:** InjectMissingFields must not invent a value an app has to have been GIVEN (a field consumed only by `after_install`), or such a field needs an "installed apps: ask" path. `audits/calibre-name-and-prune-2026-10-01/A/A2…, A3…` | **OPEN — rank P3-LOW; owner: CC (controller)** | — | — | CC |
|
||||
| **R-758** | Apps & catalog | P3 | **[P3-LOW] Eight templates declare a `mem_limit` that is not the sum of their compose limits, and no gate checks it.** FOUND 2026-10-01 by `scripts/onboarding_gaps.py` (the new-app checklist's gap page, row 5.3): adventurelog 384M vs 896M, bookstack 512M vs 768M, calcom 768M vs 1792M, claper 384M vs 640M, kimai 384M vs 640M, nextcloud 1024M vs 1664M, outline 768M vs 1152M, zipline 512M vs 768M — every one UNDER the sum, so the deploy screen and the box's capacity figure (`09` decision 22) read less memory than the app may take. REUSE.md §2 says `mem_limit` = the sum. The compose limits are what Docker enforces, so no app is starved by this; the figure the household and the capacity check read is wrong. **Needs:** the eight figures corrected (a template change: needs its own session, no image move), and a static gate (`--fast`) with a decoy, so a new app cannot repeat it. `app-catalog-felhom.eu/onboarding/EXISTING-APPS-GAPS.md` | **READY — rank P3-LOW; owner: CC (catalog)** | — | — | CC |
|
||||
| **R-762** | Apps & catalog | P3 | **[P2-MEDIUM] wger serves no CSS or JavaScript and no uploaded photo: every static file and every `/media/` file answers 404.** MEASURED 2026-10-01 on 9202 (drill catalog, the live template `82fff32`, wger 2.7), found by checklist rows 1.7 and 2.8: the login page links `/static/css/workout-manager.css`, `/static/bootstrap-compiled.css` — both 404 through traefik; the static root inside the container is empty (4 KB). A progress photo posted to `/api/v2/gallery/` answered 201 and the file is on the media volume, but `GET /media/gallery/…png` answers 404 signed in, without a session, and straight at the app inside the container. **Cause, read in the image:** the entrypoint runs `collectstatic` only when `DJANGO_DEBUG == "False"` and the template sets no `DJANGO_DEBUG`; and wger serves `/media/` only in development (`urls.py:393` „served like this during development only”) — upstream's production setup puts nginx in front for `/static` and `/media`. So the household gets an unstyled app and photos that never show. The same lines stand since the template was written (the 2026-09-29 template too). Not checked: whether any box runs wger (on 2026-09-30 none reported to the hub). **Needs:** `DJANGO_DEBUG=False` (collectstatic) and something that serves `/static` + `/media` (upstream's nginx sidecar, or the gunicorn switch of R-755 plus a static server), proven on the bench and on 9202 with a page that loads its CSS and a photo read back. Owner decides together with R-755 (same server question). `audits/new-app-checklist-2026-10-01/C/C8-signup-guest-media-static.txt`, `C/C4-seed-photo-size.txt` **-- 2026-10-01 (operator):** wger is `lifecycle: hidden` until this and its twin are fixed (catalog `55b8c8a`; read back on 9202: not on the app list, mealie control present). | **READY — rank P2-MEDIUM; owner: CC (catalog)** **Re-ranked 2026-10-03: P2→P3: wger is hidden from installs and no box runs it; needed only before it is offered again.** | — | — | CC |
|
||||
| **R-762** | Apps & catalog | P3 | **[P2-MEDIUM] wger serves no CSS or JavaScript and no uploaded photo: every static file and every `/media/` file answers 404.** MEASURED 2026-10-01 on 9202 (drill catalog, the live template `82fff32`, wger 2.7), found by checklist rows 1.7 and 2.8: the login page links `/static/css/workout-manager.css`, `/static/bootstrap-compiled.css` — both 404 through traefik; the static root inside the container is empty (4 KB). A progress photo posted to `/api/v2/gallery/` answered 201 and the file is on the media volume, but `GET /media/gallery/…png` answers 404 signed in, without a session, and straight at the app inside the container. **Cause, read in the image:** the entrypoint runs `collectstatic` only when `DJANGO_DEBUG == "False"` and the template sets no `DJANGO_DEBUG`; and wger serves `/media/` only in development (`urls.py:393` „served like this during development only”) — upstream's production setup puts nginx in front for `/static` and `/media`. So the household gets an unstyled app and photos that never show. The same lines stand since the template was written (the 2026-09-29 template too). Not checked: whether any box runs wger (on 2026-09-30 none reported to the hub). **Needs:** `DJANGO_DEBUG=False` (collectstatic) and something that serves `/static` + `/media` (upstream's nginx sidecar, or the gunicorn switch of R-755 plus a static server), proven on the bench and on 9202 with a page that loads its CSS and a photo read back. Owner decides together with R-755 (same server question). `audits/new-app-checklist-2026-10-01/C/C8-signup-guest-media-static.txt`, `C/C4-seed-photo-size.txt` **-- 2026-10-01 (operator):** wger is `lifecycle: hidden` until this and its twin are fixed (catalog `55b8c8a`; read back on 9202: not on the app list, mealie control present). **Merged 2026-10-05 from R-755 (duplicate):** `templates/wger/docker-compose.yml` still sets no `WGER_USE_GUNICORN` — the gunicorn switch is the same server question. | **READY — rank P2-MEDIUM; owner: CC (catalog)** **Re-ranked 2026-10-03: P2→P3: wger is hidden from installs and no box runs it; needed only before it is offered again.** | — | — | CC |
|
||||
| **R-774** | Apps & catalog | P3 | **[P3-LOW] Two things the new apps' pages do not show yet: Karakeep's mail-ON path is unproven, and its official phone app reports crashes to its makers.** READ/MEASURED 2026-10-01: Karakeep's `smtp_mapping` (plaintext :2526, as Cal.com) was proven only with mail OFF (a fresh install boots) — 9202 has no hub, so the relay cannot be exercised there; the official mobile app ships Sentry crash reporting with a hard-coded DSN (`apps/mobile/app/_layout.tsx`, FIT.md). **Needs:** one password-reset mail from Karakeep on a hub-enabled box (demo-hp 9201, a throwaway install); and a sentence on the page about the phone app (copy, freeze). `audits/new-apps-2026-10-01/bench/karakeep-mail-off-boot.txt` | **READY — rank P3-LOW; owner: CC (catalog)** | — | — | CC |
|
||||
| **R-76** | Apps & catalog | P4 | **FileBrowser-created folders break the setgid chain, and a drop-zone's mode is not stable** **MIGRATED FROM `ROADMAP.md` 2026-08-22 (R-369) — originally filed 2026-07-26, size S, roadmap state `idea (surfaced by the R-75 spike, 2026-07-26)`.** Moved verbatim; nothing added or reinterpreted. The roadmap keeps its copy as history, marked moved. | **OPEN — migrated from ROADMAP 2026-08-22, rank unchanged** | — | Two related findings from `audits/SPIKE-catalog-data-paths-2026-07-26.md` P3/P5, both **pre-existing** and deliberately left alone by that spike. **(a)** FileBrowser Quantum 1.3.3 creates files `0644` and folders `0755` and does **not** propagate the setgid bit — even though the entrypoint wrapper's `umask 002` really is in effect (`/proc/1/status` `Umask: 0002`). Group inheritance itself works (a file uploaded into a 2775 group-100 dir landed group 100, not the process gid 1000), so the convention's *group* half holds and only its *mode* half is lost. The consequence is proven with a control: inside a UI-created `0755` folder a gid-1000 process's file landed group **1000**, while the identical write into the 2775 parent landed group **100**. So **any folder a customer creates through FileBrowser breaks the shared-group chain one level down.** Latent today — every userdata-touching catalog app that declares an identity declares uid/gid **1000**, the same uid FileBrowser runs as, so owner permissions mask it; it bites the day a content app runs as a different non-root uid with gid 1000. The comment at `infra/infra.go:156` is right that the image ignores `-e UMASK` but does not say t | CC |
|
||||
| **R-284** | Apps & catalog | P4 | **„A kiválasztott tárhely majdnem megtelt." on a store that is 93 % FREE — an apparent inverted threshold.** Calibre-Web's deploy page rendered `<option value="/mnt/sys_drive" data-free-percent="93">` alongside „Tárhely (sys_drive) — **64.2 GB szabad**" and the warning „A kiválasztott tárhely majdnem megtelt." 93 % free read as 93 % used is the obvious candidate, and `checkStorageSpace(this)` is the function to look at. **Not confirmed by reading the code** — reported as measured output only. A capacity warning that cries wolf on an empty disk is one a customer learns to click past | **VERIFY** (2026-10-03 triage: Probably not a defect: the warning is hidden by default and shown only when free space is below 20%; the measurement likely read raw HTML without running scripts — felhom-controller/controller/internal/web/templates/deploy.html:622,808) — **READY (XS) — NEW 2026-08-09** | — | Check `checkStorageSpace`'s comparison against `data-free-percent`; add a render test per branch | CC |
|
||||
@@ -168,7 +164,6 @@ stopping line that lies.
|
||||
| **R-718** | Apps & catalog | P4 | **[P3-LOW] "Close sign-up now" restarts an app with its own switch, and the card does not say so.** MEASURED 2026-09-29 on demo-hp: pressing it on adventurelog recreated its backend (~30 s, one 500 on its login page). The window's card says the app restarts; the close card does not. **Fix direction:** the close card and the gate-open moment say "the app restarts once" where `after_setup.env` exists. **ALSO MEASURED 2026-09-29 (new-household drill):** the gate-open press on a fresh vikunja restarted it for its own switch — the front door answered 404 for ~2 s and nothing said so. | **OPEN — P3; owner: CC** **Re-ranked 2026-10-03: P3→P4: a short unannounced restart; copy polish only.** | — | — | CC |
|
||||
| **R-760** | Apps & catalog | P4 | **[P3-LOW] vikunja's compose has no healthcheck, and nothing says why.** FOUND 2026-10-01 by `scripts/onboarding_gaps.py` (row 4.1): two services in the catalog have no compose `healthcheck:` — `adventurelog-frontend` (deliberate, a comment cites R-655: the image brings its own) and `vikunja` (no comment). Not measured: whether the vikunja image declares its own `HEALTHCHECK`. The controller's probe still runs (`healthcheck.checks` in `.felhom.yml`), so the badge is not blind; Docker's own health state is. **Needs:** read the image's config; either a compose healthcheck of the family the image has (REUSE.md §2), or a comment saying why none. `app-catalog-felhom.eu/onboarding/EXISTING-APPS-GAPS.md` | **READY — rank P3-LOW; owner: CC (catalog)** **Re-ranked 2026-10-03: P3→P4: the box's own health probe still runs; only Docker's health state is blind.** | — | — | CC |
|
||||
| **R-764** | Apps & catalog | P4 | **[P3-LOW] wger sends no mail: its mail backend is the console, so a password-reset mail never leaves the box, and the template maps no SMTP.** READ 2026-10-01 inside the app on 9202 (checklist row 7.1): `EMAIL_BACKEND django.core.mail.backends.console.EmailBackend`; the settings read `ENABLE_EMAIL`, `EMAIL_HOST`, `EMAIL_PORT`, `FROM_EMAIL` …; the template carries no `smtp_mapping`. The household's admin can reset another member's password in the app; a member who forgets theirs and asks wger by e-mail gets nothing, and the page does not say so. Not measured: what wger shows after a reset request. **Needs:** an `smtp_mapping` (vaultwarden's shape, a fresh install with mail OFF booting — REUSE.md §2), or the page saying mail is not available. `audits/new-app-checklist-2026-10-01/C/C2-static-reads.txt` | **READY — rank P3-LOW; owner: CC (catalog)** **Re-ranked 2026-10-03: P3→P4: wger is hidden and no box runs it.** | — | — | CC |
|
||||
| **R-766** | Apps & catalog | P4 | **[P3-LOW] A new app's logo and screenshots reach the boxes only with the next HUB release — the website alone is not enough.** READ 2026-10-01: the box's asset syncer reads the hub's `/api/v1/assets/manifest` (`felhom-controller internal/assets/syncer.go`); the hub serves its PVC, seeded at start from `/usr/share/felhom/assets-seed/` baked into the hub IMAGE (`hub/Dockerfile:27`; the hub build copies `website/assets/*-logo.{svg,png}` and `*-screenshot-*.webp`, `hub/README.md`). So Radicale's assets (pushed `40f0742`) answer 200 on felhom.eu, and a box shows an app card without a logo until a hub build carries them. Today's fence kept the hub untouched. **Needs:** the next hub release (any) carries them — say so in its report; or a ruling that the hub reseeds from the website without a release. `audits/new-apps-2026-10-01/S/S3-assets.txt` | **READY — rank P3-LOW; owner: CC (next hub release)** **Re-ranked 2026-10-03: P3→P4: only a missing logo until the next hub release.** | — | — | CC |
|
||||
| **R-768** | Apps & catalog | P4 | **[P3-LOW] Grimoire is not built: upstream rules out public exposure and publishes no image for its current line.** READ 2026-10-01: SECURITY.md „Public-network exposure is not a supported Grimoire mode”; docs/06-remote-access.md „General REST routes remain loopback-only and tokenless”; v1.x has no published image (the compose builds from source; Docker Hub is the old 0.x). Karakeep covers the bookmark need. **Reopens if** upstream publishes an image and supports authenticated remote use. `audits/new-apps-2026-10-01/FIT.md` | **WATCHING — rank P3-LOW; owner: CC (re-read at the next catalog campaign)** **Re-ranked 2026-10-03: P3→P4: a declined app idea, watched only.** | — | — | CC |
|
||||
| **R-769** | Apps & catalog | P4 | **[P3-LOW] Pinchflat is not built: upstream is paused (no release in 2026, last push 2025-12-16) and its last release has no image tag.** READ 2026-10-01: ghcr's newest version tag `v2025.6.6`, `latest` amd64 only; runs as root by default; an unanswered 30 GB yt-dlp memory report (#866); third parties call it unmaintained (community-scripts #15968). Forks with images exist (Pinchflat-NGX, MorganKryze). **Needs:** the operator's word on a fork (a new upstream), after the YouTube sentence (R-767). `audits/new-apps-2026-10-01/FIT.md` | **WAITING-ON-OPERATOR — rank P3-LOW; owner: operator** **Re-ranked 2026-10-03: P3→P4: a new-app idea waiting on a decision.** | — | — | operator |
|
||||
| **R-770** | Apps & catalog | P4 | **[P3-LOW] Invidious — fit check only; the recommendation is not to build it.** READ 2026-10-01: playback needs `invidious-companion` (rolling `latest`, no version tags); PostgreSQL 14 (EOL 2026-11); `registration_enabled: true` by default; upstream: a bot check means „your IP is blocked from YouTube”, a 429 can last 24 h, triggered by „someone on your network” — on our boxes that IP is the household's. One bad period in 2026 (March, ~2 weeks). No report found of a family's other devices being bot-checked (inference). **Needs:** the operator's go / no-go. `audits/new-apps-2026-10-01/FIT.md` | **WAITING-ON-OPERATOR — rank P3-LOW; owner: operator** **Re-ranked 2026-10-03: P3→P4: a new-app idea waiting on a decision.** | — | — | operator |
|
||||
@@ -177,31 +172,27 @@ stopping line that lies.
|
||||
| **R-798** | Apps & catalog | P4 | **[P3-LOW] Grimmory's template sets `SWAGGER_ENABLED=false`, which v3.5.0 does not read (it reads `API_DOCS_ENABLED`, default false).** READ 2026-10-02 (`audits/family-gate-2026-10-02/B/grimmory-reads/upstream-v3.5.0.txt`). Harmless today — the API docs are off by default. **Needs:** remove the dead line or set the right name, on the next Grimmory step. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: harmless dead setting.** | — | — | CC |
|
||||
| **R-804** | Apps & catalog | P4 | **[P3-LOW] plant-it's image is gone from Docker Hub: `msdeluise/plant-it:0.10.0` answers "pull access denied … repository does not exist".** MEASURED 2026-10-02 on the bench, logged in (`audits/persistence-sweep-2026-10-02/A/sweep/batch-*.txt`, plant-it). The app is already `lifecycle: abandoned` (not offered for new installs); a box that runs it could no longer reinstall or restore it from images. **No box runs it** (read 2026-10-02). **Needs (operator):** keep the abandoned template as it is, or hide it entirely. | **WAITING-ON-OPERATOR — rank P3-LOW; owner: operator** **Re-ranked 2026-10-03: P3→P4: the app is already abandoned and no box runs it.** | — | — | operator |
|
||||
|
||||
## App updates — 19 rows (P3 12, P4 7)
|
||||
## App updates — 15 rows (P3 9, P4 6)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
| **R-440** | App updates | P3 | **[P2-MEDIUM] 23 catalog image pins float, so an update is not reproducible.** `compose pull` on a moving tag fetches whatever upstream published that day. **MEASURED 2026-09-01 over `app-catalog-felhom.eu` @ `29edad9c5bf4`: 79 `image:` lines across 53 apps, 66 distinct; 23 of those lines carry a tag with no patch version.** `postgres:16-alpine` (8 apps), `redis:7-alpine` (6), `mariadb:11.6` (2), plus one each of `postgres:15-alpine`, `postgis/postgis:16-3.5-alpine`, `mariadb:11.4`, `mariadb:12.3`, `ghcr.io/claperco/claper:2.5`, `ghcr.io/thomiceli/opengist:1.13`, `wger/server:2.6`. **A 24th is arguable and is recorded rather than rounded away:** `ghcr.io/immich-app/postgres:16-vectorchord0.4.3-pgvectors0.2.0` pins both extensions exactly but leaves the PostgreSQL patch floating. A customer pressing Frissites can therefore swap their DATABASE ENGINE build with no catalog change and no record; two boxes updated on two days end up different. **Severity MEDIUM on its own; it becomes BLOCKING the moment a pre-update copy exists**, because "what did we upgrade from and to" must be recordable and today it is not — which is also why R-440 must be read next to the digest discipline in Rule 10 of the spike. **MEASURED LIVE 2026-09-01 — the floating pins have ALREADY moved, with a passing control.** Running digests on demo-hp compared against what the registry serves for the same tag today: **`mariadb:11.4` MOVED** (`sha256:4f1d8d20...` -> `sha256:611a2fcc...`) and **`mariadb:12.3` MOVED** (`sha256:a02fe89c...` -> `sha256:dd9b303a...`), while `postgres:16-alpine`, `redis:7-alpine`, `mariadb:11.6` and `opengist:1.13` were SAME — **and both fully-pinned CONTROLS (`rommapp/romm:5.0.0`, `privatebin/pdo:2.0.5`) were SAME.** So on a box with ZERO visible drift by tag, pressing Frissites today silently swaps the DATABASE ENGINE build under `romm` and `bookstack`, with no catalog change and no record. **Compounding fact found while reading:** the recovery unit records `ImagePins` but the manifest comment says *"image NOT stored - re-pulled on restore"*, so a RESTORE of a floating-pinned app also re-pulls whatever is current — the same non-reproducibility on the recovery path. **HALF OF THE ANSWER SHIPPED 2026-09-02 (controller v0.233.0, slice 1): `app.yaml.installed_images` now records, per compose SERVICE, the reference AND the repo digest each container was actually created from — so "what did we upgrade FROM" is answerable on any box that has taken one lifecycle action since the upgrade.** What is still missing is the other half: comparing that digest against what the registry serves for the same tag TODAY, which needs a network call the render path deliberately does not make (see R-446). **The row therefore stays OPEN and its rank is unchanged** — recording a digest does not make a floating pin reproducible; it makes the drift measurable after the fact. `audits/SPIKE-app-update-2026-09-01.md` **— NIGHT 2026-09-23:** every image moved tonight (and the 21 backfilled moves) carries its resolved digest in the catalog's test record; the floating pins themselves still float until the box pulls by digest (`09` §6.4 part 6). **-- MEASURED 2026-09-30 on demo-hp 9201: a fresh install or a guarded Update renders `name:tag@sha256` from the ladder (`digest.go` `RenderWithLadderDigests`) — 9 of 20 services there run such a definition (adventurelog, docmost, paperless-ngx). The other 11 run TAG-ONLY definitions: bookstack, kimai, opengist, privatebin, romm were installed before v0.269.0 and not updated since, and `CarryDigests` (`digest.go:141`) keeps only a digest the running definition already has; bentopdf and calibre-web have no ladder at all. So a re-pull of those (a restore) takes whatever the tag serves that day. `audits/pg-last-six-2026-09-30/E/E2-digests-demo-hp.txt`.** **-- 2026-09-30 (evening):** calibre-web now has a ladder entry (`53a4a1d`), so a new install or a guarded Update renders its tested digest; bentopdf still has none. `audits/more-night-apps-2026-09-30/` | **NARROWED 2026-09-30 — floating only for (1) apps with no ladder entry and (2) apps installed before v0.269.0 until their next guarded Update; owner: CC.** **Re-ranked 2026-10-03: P2→P3: narrowed to apps with no ladder entry and pre-v0.269 installs; each guarded update fixes the latter.** | — | — | CC |
|
||||
| **R-446** | App updates | P3 | **[P2-MEDIUM] „Naprakész" can be FALSE, and the badge that says it cannot tell.** Slice 2 (controller v0.233.0, 2026-09-02) compares the RECORDED image reference per compose service against the reference the current template pins, and **queries no registry** — deliberately: a customer's box must not depend on reaching eight upstream registries to render a page (`felhom-controller/controller/internal/web/updatebadge.go`, `compareInstalledToTemplate`). **For the 23 floating pins that comparison is blind by construction:** `postgres:16-alpine`, `mariadb:11.6` and 21 others can carry an identical reference over an image that has moved. **MEASURED, not theorised — spike §5 found `mariadb:11.4` and `mariadb:12.3` had BOTH already moved upstream while two fully-pinned CONTROLS held.** So `romm` and `bookstack` on demo-hp would read „Naprakész" over a database engine build that is not the one the catalog now resolves to. **This is a KNOWN LIMITATION OF A SHIPPED FEATURE, filed the same session rather than left implicit**, and it is stated in the same words in `architecture/09-update-architecture.md` §8.1 and in the controller's `README.md`. The close is a digest comparison against the registry, which needs a network call, a cache and a failure posture — it is not a one-liner and it is not slice 2's job. **Depends on R-440**, whose fix (stop floating) would remove the problem instead of measuring it — take that route first if it is available. `architecture/09-update-architecture.md` **MEASURED 2026-09-21, and the blind spot is not one or two pins.** `audits/UPDATE-ARC-STATE-2026-09-21.md` §3.3: the catalog carries **10 floating pins of 66** (recounted — the old "23" was stale), and **6 of the 7 measurable engine pins have been repushed upstream since the catalog set them** — `postgres:16-alpine` (8 apps), `postgres:15-alpine`, `redis:7-alpine` (6 apps), `mariadb:11.4`, `mariadb:12.3`, `postgis:16-3.5-alpine`; only `mariadb:11.6` has not. The 8th (immich's own ghcr build) is UNMEASURED — ghcr exposes no anonymous last-modified timestamp. **So on demo-hp today four apps read „Naprakész" over a database engine image that has demonstrably moved.** The fix does NOT need the box to query a registry: the catalog can record each pin's digest at push time (`check-image-resolvable.py` already resolves it) and the box compares digests. Put to the operator as `09` §3b **Q6**, recommended YES — the cheapest real improvement on the arc's list. **— UPDATE NIGHT 2026-09-21:** **MEASURED ON A BOX 2026-09-21 (update night, leg B8), and it REFINES the row in two ways rather than merely confirming it.** §8.1's numbers came from a registry sweep on DooPlex; this is the same question asked of a customer-shaped box, where the badge actually renders. On guest 9202, `docmost`'s two floating pins were read as `installed_images` records them and compared against the upstream digests measured the same night: `postgres:16-alpine` → **`sha256:721873c34ceb9…` on the box and `sha256:721873c34ceb9…` upstream**, and `redis:7-alpine` → **`sha256:858f009f9709c…` both sides**. **Identical. So the badge „Naprakész" is TRUE for this box**, and the app reads correctly. **(1) The defect's size is set by INSTALL AGE, not by the catalog.** A floating pin is wrong only for a box that pulled BEFORE the tag moved; a box deployed after the repush holds the current image and its badge is right. R-446's "six repushed pins" measured the tag against the date the CATALOG set it, which is the right measure for *the catalog* and not for *a box*. **(2) The producer Q6 needs ALREADY EXISTS on the box.** `installed_images` records a real `digest` per service (`installed.go` §7.1) — the box knows exactly what it is running. What it cannot do is COMPARE, because the catalog carries no digest to compare against. That is Q6's proposal, and this is a concrete confirmation that only the catalog half is missing. Evidence: `audits/update-night-2026-09-21/23-B8-floating-pin.txt`. **-- RULED 2026-09-23 (`09` §3 decision 17):** YES — the catalog records the image digest of every pin at push time; the box compares against it and, where the catalog carries one, pulls **that exact image**, which makes a floating tag reproducible, not only the badge honest. *Pull-by-digest while the definition names a tag is a claim to verify in the build, not a ruling on mechanism.* **-- 2026-09-23:** pull-by-digest MEASURED on 9202 — Docker and Compose both pull and run `redis:7-alpine@sha256:858f…` and refuse a digest that does not exist. **Build trap, read from source:** `splitImageRef` returns "unorderable" for any ref containing `@` (`stacks/updateorder.go:134`), so a digest-carrying pin must have its digest split off before ordering or every such app reads Unknown. `09` §6.4 part 6. **— NIGHT 2026-09-23:** the CATALOG half of the close shipped: every ladder entry records the digest the registry served for Line truncated
|
||||
| **R-450** | App updates | P3 | **[P2-MEDIUM] UPDATE ARC SLICE 6 — a version sequence: automatic WITHIN a major, never ACROSS one, and an engine change gets its OWN edge.** The first half is an operator ruling of 2026-09-02 and its justification is R-449's measurement: a cross-major jump can be refused by the app itself and cannot be undone. **The second half is a rule recorded now, while it is cheap:** an engine change must never be bundled with an app version bump. `bookstack`'s `0b73e5e` moved the application 25.02.2 → 26.05.2 **and** MariaDB 11.6 → 12.3 in one commit — **two migrations behind one edge**, and an unreadable failure when it breaks. Needs a catalog-side convention and, eventually, a gate. `architecture/09-update-architecture.md` §6 **HALF SHIPPED 2026-09-21 (catalog `5ff36d098cbc`): the second half — an engine change gets its OWN edge — is now ENFORCED** by `check-engine-major.py`, which refuses a commit moving a MariaDB major together with any other image move in that template, naming what it was bundled with. The FIRST half (automatic within a major) is Slice 6 and needs four operator answers — `09` §3b **Q1–Q4**, with the shape it would take in `09` §6.2. **The urgency is now measured:** 46 of the catalog's 58 exact pins are behind upstream and **39 of those are within a major** — the population the 2026-09-02 ruling already says may move without a human. **-- RULED 2026-09-23 (operator, `09` §3 decisions 11–15):** Q1–Q4 answered. The update is a leg of the backup chain after off-site and before the full-system backup (11); automatic with a per-box switch ON by default (12); **the TEST decides, not the tag** — the box applies every step the catalog holds because the catalog holds only tested steps, and `CompareImageRefs` moves to the catalog gate (13, REPLACES "never across a major"); a box behind climbs **one tested step at a time** (14); **the box UNDOES a failed update itself** — old definition + the pre-pin safety dump + health check again, HOLD only if the undo fails (15, REPLACES §6.1's no-auto-undo). The undo and the ladder were SPIKED the same day before any build (`audits/update-rulings-2026-09-23/`); build order and costs in `09` §6.4. **-- SPIKED 2026-09-23 (`audits/update-rulings-2026-09-23/`):** the undo works by hand on three real migrating edges and needs eight product additions (R-637..R-642); **the ladder is measured absent** — one press on a box two steps behind jumped vikunja 2.3.0 → 2.5.0 in 9.5 s and 2.4.0 never ran, and the box cannot see intermediate steps at all because its catalog clone is `--depth 1` (`sync.go:283`/`:300`, one commit visible on both demo guests). The ladder's recommended format is an `update_ladder:` list in `.felhom.yml` with each intermediate step's own definition, NOT the git history (romm's image-moving commit is the definition that OOM-looped). The chain's update leg has ≤15 min as ruled (R-643). Build order `09` §6.4. **— NIGHT 2026-09-23: `09` §6.4 part 4 SHIPPED** (catalog `6db08a5`): the test record `update_ladder:` + two gates + the only writer + the 21-move backfill; 12 more steps published with records. Parts 5 (the box climbs), 6-box-half and 7 remain; the romm press on demo-hp showed today's jump live — 5.3.0 → 5.3.1 AND mariadb 11.4 → 11.8 in one press (both tested steps; `done`). **— 2026-09-24: `09` §6.4 PART 5 SHIPPED** (controller v0.268.0 `206b035`, catalog `5ed599c`): one press = one tested step, each step's own definition at `templates/<app>/steps/<StepKey(to)>.yml`, proven live on 9202 (romm 5.3.0/11.4 → 5.3.1/11.4 → 5.3.1/11.8 in two presses, `audits/ladder-2026-09-24/partD/`). Left in this row: part 6's box half, part 7 (the automatic leg), part 10 (PostgreSQL majors). **— 2026-09-25 night: part 6's box half shipped in v0.269.0/v0.269.1; PART 7 SHIPPED as controller v0.271.0** (the automatic leg, proven over six simulated nights on 9202 and watched through the demo boxes' first real night, `audits/DRILL-night-2026-09-25.md`). **Left: part 10 only (PostgreSQL majors, R-463).** **-- 2026-09-25 (evening): part 10 SHIPPED for docmost** (controller v0.273.0, catalog `afd3a60`, decisions 35–39; `audits/night-2026-09-26/`). What stays open: each of the other ten PostgreSQL apps needs its own two-venue proof before the gate lets its major move. | **NARROWED — the other ten PostgreSQL apps; owner: CC** **Re-ranked 2026-10-03: P2→P3: the gate refuses unproven PostgreSQL majors, so nothing breaks; what remains is per-app proof work.** | — | — | CC |
|
||||
| **R-440** | App updates | P3 | **[P2-MEDIUM] 23 catalog image pins float, so an update is not reproducible.** `compose pull` on a moving tag fetches whatever upstream published that day. **MEASURED 2026-09-01 over `app-catalog-felhom.eu` @ `29edad9c5bf4`: 79 `image:` lines across 53 apps, 66 distinct; 23 of those lines carry a tag with no patch version.** `postgres:16-alpine` (8 apps), `redis:7-alpine` (6), `mariadb:11.6` (2), plus one each of `postgres:15-alpine`, `postgis/postgis:16-3.5-alpine`, `mariadb:11.4`, `mariadb:12.3`, `ghcr.io/claperco/claper:2.5`, `ghcr.io/thomiceli/opengist:1.13`, `wger/server:2.6`. **A 24th is arguable and is recorded rather than rounded away:** `ghcr.io/immich-app/postgres:16-vectorchord0.4.3-pgvectors0.2.0` pins both extensions exactly but leaves the PostgreSQL patch floating. A customer pressing Frissites can therefore swap their DATABASE ENGINE build with no catalog change and no record; two boxes updated on two days end up different. **Severity MEDIUM on its own; it becomes BLOCKING the moment a pre-update copy exists**, because "what did we upgrade from and to" must be recordable and today it is not — which is also why R-440 must be read next to the digest discipline in Rule 10 of the spike. **MEASURED LIVE 2026-09-01 — the floating pins have ALREADY moved, with a passing control.** Running digests on demo-hp compared against what the registry serves for the same tag today: **`mariadb:11.4` MOVED** (`sha256:4f1d8d20...` -> `sha256:611a2fcc...`) and **`mariadb:12.3` MOVED** (`sha256:a02fe89c...` -> `sha256:dd9b303a...`), while `postgres:16-alpine`, `redis:7-alpine`, `mariadb:11.6` and `opengist:1.13` were SAME — **and both fully-pinned CONTROLS (`rommapp/romm:5.0.0`, `privatebin/pdo:2.0.5`) were SAME.** So on a box with ZERO visible drift by tag, pressing Frissites today silently swaps the DATABASE ENGINE build under `romm` and `bookstack`, with no catalog change and no record. **Compounding fact found while reading:** the recovery unit records `ImagePins` but the manifest comment says *"image NOT stored - re-pulled on restore"*, so a RESTORE of a floating-pinned app also re-pulls whatever is current — the same non-reproducibility on the recovery path. **HALF OF THE ANSWER SHIPPED 2026-09-02 (controller v0.233.0, slice 1): `app.yaml.installed_images` now records, per compose SERVICE, the reference AND the repo digest each container was actually created from — so "what did we upgrade FROM" is answerable on any box that has taken one lifecycle action since the upgrade.** What is still missing is the other half: comparing that digest against what the registry serves for the same tag TODAY, which needs a network call the render path deliberately does not make (see R-446). **The row therefore stays OPEN and its rank is unchanged** — recording a digest does not make a floating pin reproducible; it makes the drift measurable after the fact. `audits/SPIKE-app-update-2026-09-01.md` **— NIGHT 2026-09-23:** every image moved tonight (and the 21 backfilled moves) carries its resolved digest in the catalog's test record; the floating pins themselves still float until the box pulls by digest (`09` §6.4 part 6). **-- MEASURED 2026-09-30 on demo-hp 9201: a fresh install or a guarded Update renders `name:tag@sha256` from the ladder (`digest.go` `RenderWithLadderDigests`) — 9 of 20 services there run such a definition (adventurelog, docmost, paperless-ngx). The other 11 run TAG-ONLY definitions: bookstack, kimai, opengist, privatebin, romm were installed before v0.269.0 and not updated since, and `CarryDigests` (`digest.go:141`) keeps only a digest the running definition already has; bentopdf and calibre-web have no ladder at all. So a re-pull of those (a restore) takes whatever the tag serves that day. `audits/pg-last-six-2026-09-30/E/E2-digests-demo-hp.txt`.** **-- 2026-09-30 (evening):** calibre-web now has a ladder entry (`53a4a1d`), so a new install or a guarded Update renders its tested digest; bentopdf still has none. `audits/more-night-apps-2026-09-30/` **Merged 2026-10-05 from R-446 (duplicate):** the household-visible consequence — for these apps the „Naprakész" badge can never turn „behind" (`felhom-controller controller/internal/stacks/updateorder.go:96` returns false with no catalog digests / no test date); R-740 (same-tag re-test) is closed, so floating-tag drift for laddered apps is handled. | **NARROWED 2026-09-30 — floating only for (1) apps with no ladder entry and (2) apps installed before v0.269.0 until their next guarded Update; owner: CC.** **Re-ranked 2026-10-03: P2→P3: narrowed to apps with no ladder entry and pre-v0.269 installs; each guarded update fixes the latter.** | — | — | CC |
|
||||
| **R-458** | App updates | P3 | **[P3-LOW] `.felhom.yml` keeps flowing to an app whose compose file is FROZEN, so a frozen app can receive a health check written for a version it is not running.** The v0.235.0 render freezes `docker-compose.yml` for a pinned app once the catalog moves past its version, but copies `.felhom.yml` **verbatim in every case** (`Syncer.copyTemplates`). **The asymmetry is deliberate and both directions were considered:** `.felhom.yml` carries no image, and it carries `catalog_since` — the single input the update badge uses to say *„Frissítés elérhető — N napja"* — so freezing it would silently withhold the one number that tells a customer they are behind, i.e. it would break slice 2 to protect slice 3. **What it costs:** the file also carries the controller-side `healthcheck:` block and resource hints, so a template updated for a newer version can hand a frozen app a probe written for software it is not running. **THE FAILURE DIRECTION IS A FALSE ALARM, NEVER DATA LOSS** — the app keeps running; at worst it renders as degraded and, if it persisted, could reach the dead-app alarm path. That is the same class as R-330's false e-mails, which is why this is a row and not a footnote. **Not fixed now, and the reason is that the cheap fix is wrong:** freezing the whole file breaks the badge, and freezing only the `healthcheck:` key means the syncer would have to parse and re-assemble a customer-facing metadata file — new surface on the one path that touches every app on every box every 15 minutes. **What would settle it:** whether any catalog `healthcheck:` has ever been changed in the same commit as an `image:` line (measurable from the catalog's own history, no box needed). If the answer is "never", the exposure is theoretical and the row can be closed by measurement instead of by code. Owner: **CC.** `architecture/09-update-architecture.md` §5.4, §8.5 **— UPDATE NIGHT 2026-09-21:** **MEASURED 2026-09-21 (update night), leg B9, and the row's risk is NARROWER than it states.** A `.felhom.yml`-only change (a health check for a path only a newer version would serve) was pushed to a FROZEN `bentopdf` — installed `v2.8.6`, catalog ahead. §5.4's asymmetry is confirmed live: the new `.felhom.yml` reached the box while the compose `image:` line stayed `v2.8.6`. **But no false alarm was produced**: ten samples over two minutes all read `state=running` with the front door at `200`. The reason is the probe's own semantics, not luck — `healthprobe.go:258-261` treats **any response** as healthy for `type: http`, and the bogus path answers 404, which is a response. **So this row's false-alarm risk exists only for `type: api` probes carrying an `expect` block**, where the status is compared; for every `type: http` template and every `type: api` without `expect`, a newer version's path is invisible to the probe. The row's actual claim — the failure direction is a false alarm, never data loss — stands and is now measured. Evidence: `audits/update-night-2026-09-21/21-B9-frozen-app-newer-felhomyml.md`. | **READY — rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-462** | App updates | P3 | **[P2-MEDIUM] Widen the upgrade harness beyond three apps — and the cost is dominated by FIXTURES, not by machine time.** The R-449 harness works and is proven by a red negative control (`audits/SPIKE-upgrade-test-2026-09-06.md` §1). **Costed with this run's REAL numbers rather than an estimate:** a successful edge takes **6.4 s – 305.1 s, median 71.8 s**; a FAILING edge takes **556 s**, roughly **8×**, because a negative is only honest if it waits out the full settle window; 3 apps / 11 images cost **5.07 GB**, so 53 apps naively extrapolate to **~90 GB** and, at the median, about an hour of harness time for one edge each. **THAT EXTRAPOLATION UNDERSTATES THE REAL COST BY AN ORDER OF MAGNITUDE, and that is the point of this row.** Two of the three apps needed a bespoke non-browser seed route; one needed two attempts and a discarded approach; one (bookstack) can only ever be half-proven (R-460). **Fixture time scales with apps and does not amortise.** **The decision this row is really asking for is scope, not schedule:** all 53, or only the apps a customer would lose data from, or only apps whose catalog transition is a MAJOR. **Recommended shape, NOT a design — the operator picks:** start with the apps that carry a database, because §3 measured that the abort question only ever bites there. Owner: **VIKTOR rules on scope, CC implements.** `audits/SPIKE-upgrade-test-2026-09-06.md` §5 **ROW CORRECTED 2026-09-21: the scope is NOT open and this row said it was.** It read *"VIKTOR rules on scope"*; the operator ruled on 2026-09-13 (`09` §3 decision 6) that the upgrade test goes to **ALL** apps through the nightly rotation, explicitly *not* "database apps first". What is open is the WORK, not the scope. An ORDER inside that ruling — the 15 database services first, because that is where a wrong answer costs data rather than uptime — is costed as a drill brief in `09` §6.4: legs A–E ≈ **21–34 CC-hours** plus ~25–30 GB on a scratch host, with legs C (a PostgreSQL `pg_upgrade` rehearsal) and E (one automatic night on a throwaway) the two that unblock a decision. **-- UPDATE NIGHT 2026-09-21:** **The count moved from 3 apps to 21 EDGES ACROSS 19 APPS.** The update night walked real within-a-major upstream edges on scratch guest 9202 through the product's own guarded Update, each app seeded and read back through its OWN front door with a negative control on every readback: **14 proven, 3 failed, 4 inconclusive**. Proven: actualbudget, audiobookshelf, bookstack, docmost, grafana, home-assistant, mealie, n8n, navidrome, papra, privatebin, romm, vikunja, and nextcloud's MariaDB engine major. **Ten of the fourteen printed a verbatim migration line**, so the database really was rewritten and the data still read back. Box-side fixtures for 20 apps now exist at `audits/update-night-2026-09-21/fixtures.py`, and four (actualbudget, navidrome, audiobookshelf, vikunja) are ported into `app-catalog-felhom.eu/scripts/upgrade_fixtures.py` with seven new EDGES (U1-U7) so the same edges can be run on the harness venue **with their ABORT step**, which the box deliberately does not offer. **OWED, stated so it is not mistaken for done:** the U1-U7 harness RUNS (the code is in, the runs are not), and fixtures for the four inconclusive apps, of which two (vaultwarden, zipline) cannot be seeded at all while the catalog rightly closes their sign-up (see R-624). **-- 2026-09-23, the RomM lesson is IN THE HARNESS:** `upgrade-test.py` v2 runs every edge that read back under light load for `--soak` seconds (default 600) and reads the kernel's `oom_kill` counter host-side; a kill or restart turns `proven` into `failed`, a peak over 80 % of the limit adds `memory_tight`. **Red-proof:** romm 5.0.0 → 5.3.0 on the template AS PROMOTED (512M, four workers) — seeded, migrated, read back, and then **OOM-killed at +76 s** under four light callers, verdict `failed` (Docker's own OOMKilled read true here; restarts stayed 0, which is why the walk never saw it). Ten minutes is ample for this failure; demo-hp's first kill came at two hours only because nothing was loading it. Evidence `audits/update-rulings-2026-09-23/harness/`. **— NIGHT 2026-09-23:** the bench and the box now share ONE fixture set — the box walk's fixtures run on the bench through `upgrade_boxport.py` (ported verbatim into the catalog), plus a new wishlist fixture and fixes for opengist 1.15, komga (`/api/v2/users/me`) and nextcloud (wait for `occ status`). 14 apps / 15 edges tried; 12 proven on both venues and published with their test records (`audits/DRILL-night-2026-09-23.md` Part C). **-- 2026-09-30 (day brief): three new front-door fixtures (sparkyfitness, rallly, outline; catalog `e6f3ec2`) and five apps moved on both venues — the three PostgreSQL conversions (R-463) plus bookstack 26.05.5 → 26.09.1 and kimai apache-2.57.0 → 2.67.0 (upstream dropped the `apache-` prefix; the plain tag's digest equals `apache`'s,Line truncated
|
||||
| **R-469** | App updates | P3 | **[P3-LOW] REMOVE THE ENGINE-MAJOR RULE when Slice 4 (R-448) ships — a tracked act, not a lapse.** Since 2026-09-13 `app-catalog-felhom.eu` `CLAUDE.md` rules that *until the Update button takes a verified backup as its precondition, no template may move a database-engine image across a major version* (four MariaDB, eleven PostgreSQL services), and `scripts/check-engine-major.py` (fourth row of `catalog_gates.py`, run by `.githooks/pre-push` with the push range) refuses one, naming the rule and this expiry. **Why the rule:** every `mariadb:` sidecar now carries `MARIADB_AUTO_UPGRADE=1` (R-459), so a MariaDB major move CONVERTS the customer's datadir on the next Update; PostgreSQL converts nothing and refuses to start (R-463). Either way a customer-data event with no backup in front of it. **Honest limit, not re-filed:** the gate needs a parent commit and CI fetches at `--depth 1` — the R-452 gap — so on a shallow clone the runner skips it out loud and only the hook bites. **When R-448 ships:** delete the CLAUDE.md rule, the gate's row and the gate, in one commit that cites this row; then close this. **2026-09-13 — UNBLOCKED, NOT LIFTED.** R-448 shipped in controller v0.237.0/v0.238.0 (slice 4): an update now refuses without a restorable, proven Tier-2 copy, backs up first when it is stale, takes a safety dump, and holds an app that does not come up — the precondition this rule was waiting for. **The rule stays in force until someone deliberately removes it**, which is a separate act (and is worth weighing against R-475: an app with no Tier-2 copy cannot be updated at all, so the guard does not yet cover every app a major engine move would touch). **HALF-LIFTED 2026-09-21, catalog `5ff36d098cbc`.** Slice 4 shipped 2026-09-13, so the rule's own expiry condition is met — **for MariaDB**: the four `mariadb:` sidecars have both halves they need, a verified backup in front of the Update (any tier since v0.239.0) and `MARIADB_AUTO_UPGRADE=1` whose conversion the harness WATCHED run on E3/E3b with the seeded data read back after. **PostgreSQL and MySQL stay refused** — postgres performs no `pg_upgrade` and REFUSES to start on an older major's datadir across eleven templates (R-463); a backup is a route BACK, not a conversion. The refusal text now cites R-463 instead of the shipped R-448. **R-450's second half is enforced in its place:** a MariaDB major must be the ONLY image move in its template in that commit (the bookstack `0b73e5e` shape — two migrations behind one edge). The gate now PRINTS what it allowed, by name — a lifted rule that goes quiet is a lifted rule nobody can audit. Two new decoy cases; two red-proofs, each seen to fail; 40 cases green. **What remains of this row:** the PostgreSQL half, which is R-463's to clear — see `09` §3b **Q5**. **-- NARROWED 2026-09-25 (evening), catalog `6a4a5f0`, `09` §3 decision 35:** the PostgreSQL half now passes ONE app at a time — only a template whose ladder entry for the step is proven on BOTH venues and carries `engine_conversion` (the box converts it, controller v0.273.0), as the only image move in its commit. Every other PostgreSQL app stays refused; the postgis family is judged now (it was not). CLAUDE.md rule text updated the same commit. Decoys + red-proof: `audits/night-2026-09-26/C/`. | **NARROWED** — **PARTLY CLOSED 2026-09-21 — MariaDB lifted; the PostgreSQL half stands until R-463** | — | — | CC |
|
||||
| **R-607** | App updates | P3 | **[P3-LOW] A forced catalog sync answers „nincs változás" while the cache DOES change, and `catalog_images` stays stale until a separate rescan — so the update badge can be wrong for a window nobody bounds.** MEASURED 2026-09-21 on scratch guest 9202 (controller v0.260.0) while proving R-524. A real catalog move was pushed, `POST /api/sync` was invoked, and it answered **„nincs változás"** — yet the box's own cache file `<data>/catalog-cache/templates/uptime-kuma/docker-compose.yml` **had moved to the new tag**. `Stack.CatalogImages` as served by the API stayed at the OLD value until a separate `POST /api/stacks/rescan`. **WHY IT MATTERS AND WHY IT IS NOT COSMETIC:** `CatalogImages` is the single input `stacks.CatalogOrder` compares against (v0.260.0), so for that window the badge answers from a stale catalog — it can read „Naprakész" on an app that IS behind, which is the exact failure §5.6 of `09` was written to prevent, arriving by a different door. **It also cost a measurement:** the session that found it nearly recorded a `tag-ok` badge as proof of the R-524 ahead arm when the badge was in fact stale; the honest reading came only after the rescan. **An instrument that can report an old value as a current one is not a measurement.** **TWO SEPARATE QUESTIONS, and the row does not conflate them:** (a) why the sync REPORTS no change when the working tree moved — a wrong sentence, possibly a comparison against the wrong ref; (b) whether `CatalogImages` is refreshed by the sync at all or only by `ScanStacks` on its own timer — if the latter, the staleness window is the scan interval and is bounded but unstated. **Neither was isolated** — this row records the observation, not a diagnosis. **Also observed in the same run, NOT diagnosed and folded in here rather than filed twice:** a removed app leaves `applied-compose.yml` behind in its stack directory. Stated as observed; it was not established whether that is intended. **Fix shape:** first reproduce with a loop that pushes a tag, syncs, and reads `catalog_images` on a timer, so the window is a NUMBER before anything is changed. Evidence: `audits/update-arc-2026-09-21/06-sync-after-bump.txt` and `16-r524-sync-box-ahead.txt`. **— UPDATE NIGHT 2026-09-21:** **Seen again 2026-09-21 (update night), a dozen times in one session, and for the first time with a USER-VISIBLE consequence rather than a measurement one.** On `mealie` the bump was pushed, `POST /api/sync` AND `POST /api/stacks/rescan` were both run, and the badge still read the up-to-date one (HU „Naprakesz", EN "Up to date") — the catalog had not reached `catalog_images` yet. The guarded Update was then pressed and **reported „Frissitve" after 2.1 seconds having moved nothing at all**: pinned, installed, the live compose line and `docker inspect` all still read `v3.20.1`. That is honest given a stale cache — the pin is written from *the catalog's current definition*, which was still the old one — but what the household sees is a button that says it updated them and did not. **A NUMBER, at last, which is what this row asks for:** the night's harness was changed to poll `catalog_images` until the pushed reference appears and to report how long that took; those figures are each edge's `badge_catchup_seconds`, and here they are: **4.4 s, 4.4 s, 4.5 s, 4.5 s — and 29.0 s.** The four fast ones are one sync+rescan round; the 29-second one (`nextcloud`, an engine-sidecar bump) needed **additional** sync+rescan rounds before `catalog_images` carried the pushed reference. **So the window is not a fixed scan interval — it varies by roughly 7x between edges on the same box in the same hour**, which is why a caller (or a household) cannot know when the badge is safe to read. Before tonight this row had no number at all; it now has five, and they disagree with each other, which is itself the most useful thing about them. Every drill-catalog bump of the night was followed by `POST /api/sync` answering „Sablonok naprakészek — nincs változás" while the box's cache HAD moved, with `catalog_images` staying stale until a separate `POST /api/stacks/rescan`. The night's harness therefore rescans unconditionally after every sync, which is a workaround and not a fix. **The window was still never measured as a NUMBER** — that is what the row asks for and what remains owed. **— NIGHT 2026-09-23, measured twice more:** for an INSTALLED app the stack directory's compose is the applied one, so waiting on it for a drill commit took ~7 min (chaos rounds 2 and 4); the badge's own `catalog_images` caught up in seconds. And twice a fresh box walk deployed a STALE template (the uptime-kuma "after" run, the opengist re-walk) because the deploy ran before the sync reached the file — the walks now wait for the box's template to show FROM. | **READY — rank P3-LOW; owner: CC (controller)** | — | — | CC |
|
||||
| **R-615** | App updates | P3 | **[P3-LOW] Pointing a box at a different app catalog by `git.repo_url` alone is INERT — the box keeps fetching from the repository it first cloned.** FOUND 2026-09-21 by reading `sync.go` **before** running it, which is the only reason the update night's drill catalog worked at all. `Syncer.gitCloneOrPull` (`controller/internal/sync/sync.go:274-306`) clones **only when `<data>/catalog-cache/.git` is absent**; on every later cycle it runs `git fetch --depth 1 origin <branch>` + `git reset --hard origin/<branch>` **against the remote stored in the clone**, which `buildRepoURL` wrote at clone time. Changing `git.repo_url` in `controller.yaml` and restarting therefore changes **nothing**: the sync keeps pulling the old catalog and reports success. Measured: after the repoint, `git -C <data>/catalog-cache remote -v` still read `app-catalog-felhom.eu`; the box only followed the drill repo once the cache directory was removed. **Why it matters beyond a drill:** this is the one knob that would move a box to a different or a staged catalog — for a migration, a per-customer catalog, or a rollback of the catalog itself — and it silently does not work. **Nothing is wrong with the CACHING**, which is right; what is missing is that a changed `repo_url` must invalidate the clone. **Fix shape:** on start, compare `git.repo_url` with the clone's `origin` and re-clone when they differ (or `git remote set-url` + a full fetch); log which happened. A test that changes `repo_url` under an existing cache and asserts the next sync reads the NEW repo — it fails today. Evidence: `audits/update-night-2026-09-21/04-9202-config-pre.txt`, `05-9202-follows-drill.txt`. **-- 2026-10-01 (night, new apps):** the other direction measured on 9202: after `repoint restore` a DRILL-only template (grimmory) stayed OFFERED on the app list, because its `/opt/docker/stacks/grimmory` folder (written by the drill sync) remained; removing the folder by hand ended it (`audits/new-apps-2026-10-01/B/B9-restore-live.txt`). Older drill folders (`chaoscrash`, `chaosoom`, `chaosoomb`) sit there too. The drill teardown should list and remove stack folders the live catalog does not have. | **READY — rank P3-LOW; owner: CC (controller)** | — | — | CC |
|
||||
| **R-622** | App updates | P3 | **[P2-MEDIUM] `adventurelog v0.13.0` migrates the customer's database and then does not serve — the edge must NOT be promoted, and it is the first real-catalog candidate this project has measured as unsafe.** MEASURED 2026-09-21 on guest 9202 through the product's own guarded Update. `v0.12.1 → v0.13.0` (backend AND frontend together, PostGIS held constant). The backend applied **nine migrations, every one `... OK`** — `adventures.0072_trail_wanderer_author_fields` through `integrations.0009_alter_endurainintegration_auth_method`, plus `billing.0001_initial` — and then the container's own healthcheck failed with `URLError: [Errno 111] Connection refused` on five consecutive checks. The app never bound its port. The update held honestly after the full 5-minute wait. **THE PRODUCT DID EVERYTHING RIGHT AND THAT IS HALF THE FINDING:** the precondition found a Tier-1 copy one minute old, the safety dump was written, the pin advanced BEFORE the pull, the health wait was not short-circuited, the app was stopped rather than left half-running, and the hold sentence named the tier, the date and what the copy holds — „saját meghajtó, 2026-09-21 20:47 — ez a másolat a beállításokat, az adatbázist és az adatköteteket tartalmazza." **This is exactly the case `09` §4 exists for:** the migration RAN, so there is no undo, only a restore — and the restore is the thing slice 4 made sure existed first. **What it needs:** adventurelog stays OFF the promotion list; the cause is not diagnosed here (R-621 is why); and before it is ever promoted the edge should be re-run on the harness with its ABORT step, since an app that migrates and then refuses is the shape most likely to refuse the old image too. Evidence: `audits/update-night-2026-09-21/apps/adventurelog/`. | **READY — rank P2-MEDIUM; owner: CC (catalog); NOT a Felhom defect — an upstream edge that fails** **Re-ranked 2026-10-03: P2→P3: the failing version is simply not promoted, so no household receives it.** | — | — | CC |
|
||||
| **R-683** | App updates | P3 | **[P3-LOW] Watch: after a power cut during an update's health check, the hold named an HOUR-OLD second-drive copy, not the one the update's own backup should have just made.** 2026-09-24 chaos round 3 (nextcloud, `backup_max_age: 1m`): no `backing-up` phase was seen and the hold named Tier 2 at 13:04 for an update pressed at 14:04; the pre-cut controller log was lost with the container (the runner now saves it at arm time — R-320). Round 11, the same action without a power cut, named a fresh 14:34 copy and logged the Tier-2 copy. The sentence was TRUE (it named the copy it offered); the question is why the update did not back up first. Not reproduced; watch the next power-cut drill. `audits/night-2026-09-24/E/round-03*.json`, `E/round-11-controller-pre.log` | **OPEN — P3; owner: CC (watch)** | — | — | CC |
|
||||
| **R-738** | App updates | P3 | **[P2-MEDIUM] Every wger update that brings database migrations leaves wger broken, and the guarded Update reports `done`.** MEASURED 2026-09-30 on scratch guest 9202 (drill catalog): wger 2.6 → 2.7 through the product's guarded Update ended `done` in 72.8 s (health: the front page answers); the web login then answered **500** (`no such column: core_userprofile.time_zone`), still 500 a minute later, **12 migrations unapplied** (`manage.py showmigrations`). Cause, read from the image: `entrypoint.sh` runs `manage.py migrate` only when `DJANGO_PERFORM_MIGRATIONS=True`; the template does not set it (a fresh install works because `wger bootstrap` builds a new database). The harness caught it: the box verdict is `failed`, and no ladder entry was written. No box reporting to the hub runs wger (hub `/apps`, same day). **Also a product gap:** the guarded Update's health check cannot see an app whose front page serves while its data is unusable — the fixture's read-back is what saw it. **Fix:** `DJANGO_PERFORM_MIGRATIONS=True` in the template (the image's own switch for it), then the step again on both venues. `audits/more-night-apps-2026-09-30/box/wger/` **-- FIXED 2026-09-30:** catalog `7a4ff48` sets `DJANGO_PERFORM_MIGRATIONS=True`; the same 2.6 → 2.7 step then ran the migrations (`Applying …` in the log) and read back on 9202 (46.2 s) and on the bench (proven, 0 kills); published `4ad32aa`. The product-side gap (a health check that sees only the front page) remains, as a note: the fixture's read-back caught it, the box could not. | **NARROWED** (2026-10-03 triage: the row's verdict was finished, but it names open work no other row carries — the guarded Update's health check reads only the front page, so it cannot see an app whose front page serves while its data is unusable) — **CLOSED 2026-09-30 — catalog `7a4ff48` + `4ad32aa`** | — | — | CC |
|
||||
| **R-785** | App updates | P3 | **[P3-LOW] SparkyFitness is pinned 11 releases and a major behind upstream (v0.17.3; upstream v1.7.3, v1.6.0 dated 2026-07-24).** READ 2026-10-01 (`audits/visitors-2026-10-01/C/bench/C1-previous-tag.txt`). **Needs:** an update walk 0.17 → 1.x through the ladder (bench + box), after R-784 is decided. | **OPEN — rank P3-LOW; owner: CC (after R-784)** | — | — | CC |
|
||||
| **R-460** | App updates | P4 | **[P3-LOW] BookStack's FILE half cannot be seeded or verified without a browser, so its upgrades can only ever be auto-proven for the DATABASE.** MEASURED 2026-09-06 while building the R-449 harness. BookStack's API needs a token that is only mintable through its web UI, and its HTTP login is unusable headlessly for a second, independent reason: `APP_URL` comes from the template as `https://${SUBDOMAIN}.${DOMAIN}`, so the app marks its session and XSRF cookies **`secure`**; curl over plain http stores neither and **every login POST returns 419 Page Expired**, which looks exactly like a wrong password. The container serves no TLS. **The database half IS provable** — the harness seeds with `php artisan bookstack:create-admin` and reads back with a DIFFERENT artisan command that must find the record, carrying its own negative control on every call. **What is unprovable is an uploaded image or attachment**, i.e. exactly the half a customer would notice. **THIS IS A FACT ABOUT THE APP, NOT A DEFECT IN THE HARNESS**, and it is recorded because Slice 6 needs to know which apps can be auto-verified and which can only be partly verified — nobody had that list before. **Deliberately NOT worked around:** planting a file in the volume would make the test pass while proving nothing, which is R-156's exact failure. **What would remove it:** a headless token route (upstream), or accepting a browser-driven step for this app alone, which DooPlex cannot run. Owner: **CC.** `audits/SPIKE-upgrade-test-2026-09-06.md` §6 **-- UPDATE NIGHT 2026-09-21:** bookstack's edge was walked again on 2026-09-21 and is again **half-proven**: the database half read back through `php artisan` with its own negative control, the file half untouched. The limitation is unchanged and is now measured on the box as well as on the harness. Two more apps joined the same class tonight for a different reason (R-624). | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: a limit of test coverage for one app, not a defect a household meets.** | — | — | CC |
|
||||
| **R-464** | App updates | P4 | **[P3-LOW] MariaDB's entrypoint prints `MariaDB upgrade not required` on an UNSUPPORTED DOWNGRADE, so that line cannot be used as a soundness signal.** MEASURED 2026-09-06. After converting a datadir to `12.3.3-MariaDB` and then starting **11.6** on it, the entrypoint logs, on every start: **`[Note] [Entrypoint]: MariaDB upgrade not required`**. Asked properly, the same engine answers **`FATAL ERROR: Version mismatch (12.3.3-MariaDB -> 11.6.2-MariaDB): Trying to downgrade from a higher to lower version is not supported!`** **The entrypoint compares the datadir's recorded version against its own and concludes there is nothing to DO. That is true, and it is not a statement that the state is sound.** **THIS IS THIS PROJECT'S MOST-REPEATED CLASS, in a new costume** — the same shape as `CLAUDE.md`'s "presence is not success" and as R-443's HTTP 200 over a crash-looping app: a reassuring sentence that answers a narrower question than the one a reader will take it for. **Why it is worth a row rather than a footnote: the obvious cheap instrument for R-459 is to grep container logs for that exact line**, and such an instrument would report "fine" for an unsupported downgrade. **The correct probe is `mariadb-upgrade --check-if-upgrade-is-needed`**, which is what `upgrade-test.py`'s `engine_state_after` now uses. **Also recorded, because it nearly produced a wrong answer here: run without credentials that command returns `ERROR 1045 … FATAL ERROR: Upgrade failed` with exit 1** — an authentication failure wearing the shape of a verdict. Owner: **CC.** `audits/SPIKE-r459-mariadb-upgrade-2026-09-06.md` §5.4 | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: a recorded lesson; the harness already uses the correct check.** | — | — | CC |
|
||||
| **R-610** | App updates | P4 | **[P2-MEDIUM] The power cut AFTER the new version has started was never measured — R-520 closed on the safe half only.** R-520 (CLOSED 2026-09-21) cut in `pulling`, where **nothing had run**: the pin goes back and that is the easy case. Its own last lines said the `starting` cut was "NOT measured and does not re-open this row", and no row carried it. The dangerous case is the cut AFTER the new version has started and may already have migrated the customer's data — where `RecoverUpdates` (`stacks/update.go:909`) marks the app Updating and RESUMES rather than rolling back. **That behaviour was READ from the source and never observed.** **CLOSED 2026-09-21 — measured THREE times on guest 9202, controller v0.260.0**, with three different apps and two different cut mechanisms: vikunja 2.3.0→2.6.0 and uptime-kuma 2.4.0→2.5.0 by `pct stop` (a real power cut), and wishlist v0.66.0→v0.67.0 by restarting ONLY the controller container (exactly what a self-update does). **All three ended HONEST:** the recovery line appeared, the update resumed, each app came up on the NEW version, and in every case all FOUR version observables agreed — `pinned_images`, `installed_images`, the live compose `image:` line, and `docker inspect` of the running container (digests matched too). No hold, no stuck `Updating`, no surviving journal, no retry loop. The seeded data read back through each app's own front door for the two where a post-cut read-back was taken. **THE DANGEROUS CASE WAS GENUINELY EXERCISED, and the proof is a log line, not an assumption:** vikunja's own log shows `Ran all migrations successfully` and `Vikunja version v2.6.0` at **12:28:26.881 UTC — 0.64 s after the cut decision and ~0.4 s before the guest stopped answering.** The 2.6.0 schema migration had ALREADY been applied to the customer's SQLite database when the power went. Recovery resumed FORWARD, so old-binary-on-migrated-database never happened — **but this branch is one step from it: had the cut landed a second earlier, in `pinning` or `pulling`, the 2.3.0 pin would have been put back onto a 2.6.0 database.** That is not a defect today; it is the reason §4's "no automatic rollback" ruling is right, and it is now evidence rather than argument. **INSTRUMENT LIMIT, stated because it bounds the claim:** `starting` lasts well under a second on this box. Three attempts across two cut mechanisms (`pct stop` returning in 3.0–3.8 s; a controller restart in 1.7 s) ALL landed in `verifying`. No phase was faked. **`RecoverUpdates` handles `starting` and `verifying` in ONE branch, so all three runs exercise the same recovery arm** — the arm under test. A cut that lands inside `starting` itself remains unmeasured and would need an in-process fault injector. Evidence: `audits/update-arc-gaps-2026-09-21/` 04, 05, 07. | **NARROWED** (2026-10-03 triage: the row's verdict was finished, but it names open work no other row carries — a power cut landing INSIDE the `starting` phase is unmeasured (all three cuts landed in `verifying`); needs an in-process fault injector) — **CLOSED 2026-09-21 — measured three times** | — | — | CC |
|
||||
| **R-618** | App updates | P4 | **[P1-HIGH] THREE apps are presented to the household as UNHEALTHY while they are working perfectly — and because the guarded Update waits on that same probe, a SUCCESSFUL update ends by STOPPING the working app and sending the household to a restore they do not need.** **RANK RAISED FROM P2 TO P1 BY A LIVE MEASUREMENT taken the same night, and the escalation is the whole point:** tandoor's Update 2.6.13 → 2.6.15 was pressed at 21:16:47 and entered `verifying` at 21:17:46. At **21:18:28** the NEW version was `Up 25 seconds` and answering **HTTP 200** on `/accounts/login/` through the household's own front door — while the controller, probing port 8080 where nothing listens, could not see it. `verifying` therefore cannot pass, the full `update.health_timeout` is spent, `Manager.failAndHold` runs `compose down`, and the app is STOPPED. **Nothing is lost** — the data is in the volumes and the restore works — **but one wrong port number in a template converts every successful update of that app into an outage plus an unnecessary restore, for every household running it.** Evidence: `audits/update-night-2026-09-21/14-tandoor-serving-while-verifying.txt`. MEASURED 2026-09-21 on guest 9202 (controller v0.261.0, catalog `f5f6a152b513`). Two shapes, one class: **(a) `tandoor` — the WRONG PORT.** `.felhom.yml` probes `port: 8080`; the container listens on **80 and nothing else** (`ss -ltn` inside it), the compose's own traefik label routes to 80, its own docker healthcheck reads `healthy`, and `/accounts/login/` answers **200** through the household's real front door. `GET /api/stacks/tandoor` nevertheless reads `state: "unhealthy"`. **(b) `zipline` — the WRONG PATH.** `.felhom.yml` probes `/api/health`, which zipline 4.6.1 answers **404 `Route GET:/api/health not found`**; **the compose healthcheck in the very same file uses `/api/healthcheck` and is correct and green.** `/dashboard` answers 200. The controller reads `unhealthy`. **This is the MIRROR of R-613** — that is a probe that passes on a broken app (a false GREEN, which no alarm catches); this is a probe that fails on a working app (a false RED). **IT DOES NOT ALARM, AND THAT SETS THE RANK:** `08-alarm-ladder.md` §4 puts `unhealthy` deliberately in the NOT-down set, so no dead-app event and no customer mail follows — the damage is what the household READS, plus anything that gates on `state`. **IT ALREADY COST A MEASUREMENT TONIGHT:** this drill's harness waited for `state == "running"` and hung for its full budget on tandoor, an app that was up the whole time. An instrument waiting for a wrong answer looks exactly like a slow app. **THE GATE THIS WANTS IS CHEAP AND STATIC, AND THAT IS THE FINDING'S REAL VALUE.** Both halves of the answer live in the same template: compare the `.felhom.yml` probe's port and path against the compose's **own** `healthcheck: test:` URL. A sweep of all 53 templates on that rule was run tonight and returns **five** disagreements: `tandoor` (PORT — **CONFIRMED live**), `zipline` (PATH — **CONFIRMED live**), `wger` (PORT, probe 80 vs compose 8000 — **CONFIRMED live the same night**), `home-assistant` (PATH, `/api/` vs `/manifest.json` — **NOT MEASURED**), and `adventurelog` (a FALSE POSITIVE of the sweep's own regex — it reads `running` live). **So the rule finds both real defects, with two candidates and one false positive out of 53** — a good enough signal for a fast gate, provided it reports candidates rather than convictions and a person or a runtime check resolves them. The earlier, cruder rule (probe port vs the *traefik* port) is strictly worse: it clears zipline and convicts adventurelog. **AND A SECOND FIX SHAPE, ON THE CONTROLLER SIDE, WORTH CONSIDERING BESIDE THE CATALOG ONE:** in both confirmed cases the container's OWN docker healthcheck was **green** the whole time. A `verifying` phase that is about to stop a working app could ask that too — if the compose declares a healthcheck and docker reports `healthy`, the app is alive whatever our probe thinks. That does not excuse a wrong probe, but it turns this failure direction from an outage into a wrong label. It is a design question, not a defect, and is raised here rather than decided. **Needs:** fix tandoor's port (80), zipline's path (`/api/healthcheck`) and wger's port (8000) — **all three are now CONFIRMED live, none is a guess**; add the static gate with a decoy each way (R-421) — a template whose probe agrees must not read as a disagreement, and vice versa. **THE GATE'S RULE WAS THEN SHARPENED BY READING `healthprobe.go` RATHER THAN ASSUMING IT, and the sharpening REMOVED a false conviction.** `type: http` treats **any** response as healthy (`healthprobe.go:258-261`), and `type: api` with **no** `expect` block does the same (`:265-268`); only `type: api` WITH `expect.status` cares about the path or the code. So a PATH difference is a candidate only for the third shape, while a PORT difference isLine truncated
|
||||
| **R-621** | App updates | P4 | **[P2-MEDIUM] A held update DESTROYS the evidence of why it failed: `failAndHold` runs `compose down`, the failing containers are removed, and their output is gone before anyone — household, operator or the next session — can read it.** MEASURED 2026-09-21 on guest 9202 on a REAL upstream edge: `adventurelog v0.12.1 → v0.13.0`. The new backend applied **nine Django migrations successfully** and then never listened; the update held after the full 5-minute health wait. **`Manager.failAndHold` (`stacks/update.go:723`) calls `updateCompose(dir, env, "down")`**, which removes the containers rather than stopping them, and nothing captures their logs first. Within seconds the box's own log recorded `Logs result for adventurelog: 0 bytes returned (empty)` and `docker ps -a` held nothing at all. **What survives is the WHAT and not the WHY:** the controller line `update adventurelog FAILED after the new version was started: not healthy: not healthy within 5m0s (last: state unhealthy)` and the household's sentence, both of which say the app did not come up and neither of which says the migrations ran and the server then failed to bind. **This is R-320 ("evidence off the machine before the teardown") as a PRODUCT behaviour rather than a session habit** — the teardown here is the product's own, it is correct to perform (a half-started new version must not keep running), and it happens before anyone can look. **Why it matters beyond a drill:** the hold sentence sends the household to a restore, and after the restore the only remaining question is *should I press Update again?* — which nobody can answer, because the one artefact that would say so no longer exists. It also makes every future held update unreportable to an upstream project. **Fix shape:** capture `compose logs --no-color --tail N` into the stack directory (beside `applied-compose.yml`, which already travels with the stack) IMMEDIATELY before the `down`, and surface it on the app page's hold panel or at least through the existing `/api/stacks/<n>/logs` fallback. Bounded size, written once per hold. A test that holds an app and asserts the captured file is non-empty — it fails today. Evidence: `audits/update-night-2026-09-21/apps/adventurelog/why-it-failed.txt`, `state-after-hold.txt`. **FIXED in v0.262.0.** `failAndHold` now writes each service's log into `<stackdir>/hold-logs/<ts>/compose-logs.txt` (`compose logs --no-color --tail 400`) **before** the `down` that destroys them, best-effort by design: a hold must never fail because its evidence could not be written. This is what `adventurelog` cost — nine migrations ran, the app never bound its port, and the only log that could have said why was gone before anyone looked. | **VERIFY** (2026-10-03 triage: the row's verdict was finished, but it names open work no other row carries — the fix shape also said to show the captured hold log on the app page's hold panel; the row does not say that was done) — **CLOSED 2026-09-22 — v0.262.0: the hold keeps the app's own log before stopping it** | — | — | CC |
|
||||
| **R-687** | App updates | P4 | **[P3-LOW] Part 7's live proof has four gaps a scratch box cannot close, and one observability gap.** (1) **W+5h reached with steps left** is proven by unit test only (`TestLeg_NoStepAtOrAfterW5h`) — the leg starts at W+105m and would need a 3-hour leg live; (2) **the off-site leg FAILING** before the update leg: 9202 has no off-site tier, so only the "no target" path ran live — failure and panic are `TestChainUpdateLeg_EveryPath`; (3) **a `files_may_change` step WITHOUT a whole copy**: both drill apps given the mark (wishlist, navidrome, romm) turned out whole on 9202 by the backup side's truth table (why, per app, is not logged — see the observability gap), so only "with a copy" ran live; (4) **the full-system gate waiting** cannot run on 9202 (no agent), and **did not occur on the demo boxes' real night either** (2026-09-25: both legs ended by 04:19, before the gate opened at 04:30, and no whole-box backup was due on either box) — unit + red-proof only (`TestD20_GateWaitsForTheLeg`). **Also cosmetic:** a leg with no steps reports `"steps": null` to the hub, not `[]`. **Observability:** when the leg TAKES a `files_may_change` step it does not log which whole copy allowed it (only the skip says why). `audits/night-2026-09-25/C/` **-- NARROWED 2026-09-25 (controller v0.273.0):** the cosmetic `"steps": null` → `[]` and the taken `files_may_change` step's missing log line are FIXED (red-proofed, `audits/night-2026-09-26/F/`). Items (1)–(4) stay; (4) did not occur on 2026-09-25 either (demo-felhom's whole-box backup ran at 07:29, three hours after its leg; demo-hp had none due). **-- 2026-09-28 (night 27/28):** (4) did not occur again — on demo-hp the leg ended 04:23:54 and the whole-guest backup began 04:37:06, after the gate opened at 04:30; demo-felhom's backup ran at 07:36 (`audits/evidence-golden-0276-2026-09-28/phaseD2-night-read.txt`). **-- 2026-09-30 (by day, demo-hp 9201): item (4) PROVEN LIVE.** The night chain pressed by hand, the window moved to W = now − 2h05m the moment the leg started, `quiesce.poll_interval` 1m: `[quiesce] full-system backup due and inside its window, but the automatic update leg is running … deferring` at 11:35:11 and 11:36:11 UTC while bookstack (55.1 s) and kimai (75.1 s) stepped; the leg's end line at 11:36:29; the backup quiesced at 11:37:11 (the first poll after), job done 11:47:19, the agent's `backup: completed` 9.98 GB. Config and window put back and read back (`audits/pg-last-six-2026-09-30/C/`). **Found, cosmetic, manual chain only:** the deferral names the moved window's W+5h (16:29) while the manual leg's own deadline was its start + the leg length (16:49). | **OPEN — P3, gaps (1)–(3) + the manual-chain deferral text; item (4) proven live 2026-09-30; owner: CC** **Re-ranked 2026-10-03: P3→P4: the gaps are covered by unit tests; left is live-proof completeness and one log text.** | — | — | CC |
|
||||
| **R-734** | App updates | P4 | **[P3-LOW] The harness marks immich `files_may_change` because immich rewrites six 13-byte `.immich` folder markers at every start.** MEASURED 2026-09-30 on the bench (v3.2.2 → v3.2.4): the bind-tree hash of `appdata/immich` changed; the only changed files were `{encoded-video,library,backups,profile,thumbs,upload}/.immich`, rewritten at each start — no household file. The mark is honest by the harness's rule and the ladder writer copies it (never edited by hand), so immich's v3.2.4 night step needs a fresh WHOLE copy (decision 13); on a box without one the night leg skips it and a person presses. The 2026-09-23 immich entry did not carry it (`files_changed []`). **Needs:** a decision whether app-owned marker files are excluded from the file hash (a per-template ignore list, or a size/name rule), or the mark stays. **-- 2026-09-30 (evening):** the harness now NAMES the files behind the mark (`files_changed_detail`, catalog `5b1972b`); on immich's step `0b82…` re-proof it named exactly the six `.immich` markers again. calibre-web's step (v4.0.6 → v4.0.8) carries the mark too, and the named files are its LIBRARY DATABASE: `media/books/metadata.db`, `metadata.db-shm`, `metadata.db-wal` changed; the book file did not (bench names re-run, `A/calibre-names/`). That is household data (the template's backup class `mandatory` holds DB and books as one unit), so the mark is right there and the night leg takes that step only with a fresh whole copy. On 9202 the same step changed no file in that folder (per-file hashes before/after) — not explained. `audits/more-night-apps-2026-09-30/` | **READY — rank P3-LOW; owner: CC (harness); the rule change needs a word** **Re-ranked 2026-10-03: P3→P4: the effect is an update that waits for a press; no data risk.** | — | — | CC + operator |
|
||||
|
||||
## Backup & restore — 53 rows (P2 8, P3 23, P4 22)
|
||||
## Backup & restore — 52 rows (P2 8, P3 22, P4 22)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
@@ -215,7 +206,6 @@ stopping line that lies.
|
||||
| **R-822** | Backup & restore | P2 | **An add-only key does not make retention safe: an attacker who can only ADD snapshots can make the honest pruner erase every real one.** MEASURED 2026-10-03 (lab, restic 0.14.0 from the controller image, rclone `--append-only`): 13 empty snapshots dated in the future with the same host and tag, added through the add-only key (all allowed), make the box's exact policy `forget --group-by host,tags --keep-daily 7 --keep-weekly 4 --keep-monthly 6` keep only the fakes and select **all 3 real snapshots** for removal (`--dry-run`). **Whoever prunes an append-only repo — the box in a window, or a Felhom-side worker — inherits this.** Not a defect today (today the box can simply delete, R-95); a PRECONDITION on the R-95 build, like R-430 was. `audits/offsite-append-only-2026-10-03/lab/C3-retention-poisoning.txt` | **NARROWED 2026-10-03 — the guard ships in controller v0.289.0 (future-dated / newer-than-hub / recent-removal refusals, oldest-first cap, the lab's 13-fake shape refused in a test). RESIDUAL, not closable by a guard: an add-only attacker can plant PAST-dated snapshots interleaved with real ones and so steer weekly/monthly keeps; bounded per window by `MaxRemove` and the hub's count check, not prevented.** | — | Before any `forget`: refuse when a snapshot is dated in the future or newer than the newest the hub has seen reported; dry-run first and abort above the count the policy can remove in a week; hub compares the count before/after (DESIGN.md §3) | CC |
|
||||
| **R-49** | Backup & restore | P3 | **[P2] The offsite capture set is ~90% cache and duplication — 1.1 GB of a 1.2 GB immich "photo backup".** Measured 2026-07-19: `immich_ml_cache.tar` **823 660 032 B (~60%)** — re-downloadable ML model weights; `immich_postgres_data.tar` **308 251 136 B (~23%)** — a raw tar of the postgres data dir that DUPLICATES the logical `.sql` dump captured beside it; `upload/backups/` **18 MB** — immich's own nightly dump, a backup inside the backup, growing daily; plus the stranded pre-v3 `dccc13fe…` tree (~36 MB) no DB has ever referenced. Actual irreplaceable content: **72 MB of originals**. **MIGRATED FROM `ROADMAP.md` 2026-08-22 (R-369) — originally filed 2026-07-19, size S–M, roadmap state `idea`.** Moved verbatim; nothing added or reinterpreted. The roadmap keeps its copy as history, marked moved. | **OPEN — migrated from ROADMAP 2026-08-22, rank unchanged** **Re-ranked 2026-10-03: P2→P3: wasted space and transfer, no data risk; needs a capture-set ruling, not a sale blocker.** | — | **Evidence: `audits/DIAG-immich-restore-round2-2026-07-19.md` §4 (full byte breakdown).** This is the customer's offsite quota and transfer cost, and it lands on the Hetzner sub-account they are billed for. **Recorded, deliberately not changed** — a capture-set exclusion is a data-loss-shaped decision and gets its own ruling, not a drive-by edit. Candidates in priority order: (a) `immich_ml_cache` — pure cache, strongest case; (b) the `postgres_data` volume tar where a logical dump of the same DB is already captured (the dump is what the restore path actually replays); (c) `upload/backups/`. Likely generalises past immich into a template-classification rule about cache volumes and self-backup directories, so it should be specified against the catalog, not one app | CC |
|
||||
| **R-127** | Backup & restore | P3 | **The catalog's `data_key: true` flag is UNRELIABLE — at least four data-encrypting keys the catalog itself labels as encryption keys are unflagged; and the O4 restore path can regenerate a DB password that then does not match the restored data directory** | **READY (S/M)** | — | **Found by D5's Part 0, and it is why D5's boundary is `type: secret` rather than `data_key`.** Two separable legs. **(a) The misclassification.** Only 5 fields across 4 apps set `data_key: true` (`adventurelog/SECRET_KEY`, `homebox/HBOX_AUTH_API_KEY_PEPPER`, `papra/AUTH_SECRET`, `sparkyfitness/{API_ENCRYPTION_KEY,BETTER_AUTH_SECRET}`), yet `n8n/N8N_ENCRYPTION_KEY` („Titkosítási kulcs"), `wanderer/POCKETBASE_ENCRYPTION_KEY` („Adatbázis titkosítási kulcs"), `calcom/CALENDSO_ENCRYPTION_KEY` and `bookstack/APP_KEY` are unflagged — the catalog's own Hungarian labels contradict the flag. **D5 makes this non-urgent but not harmless:** everything `type: secret` now travels, so the keys DO reach the drive; what stays wrong is the **fail-closed gate**, which only refuses for `data_key` names — so if one of these is missing from both sources the restore proceeds onto data it cannot decrypt instead of refusing. Fix = flag them (`app-catalog-felhom.eu`, a catalog-only change) + a gate/test that the flag set and the label set agree. **(b) The regenerated-DB-password trap.** `internal/backup/restore_unit.go` O4 generates a replacement for any missing non-data-key secret. Proven on `postgres:16-alpine`: with PGDATA restored from the volume tar, `POSTGRES_PASSWORD` is ignored (initdb skipped), so the app fails over the compose network while the dump replay still succeeds through the container's local **trust** socket — success reported, data unreachable. v0.188.0 corrected the WARN's false claim that *"stored data is unaffected"* and scoped it, but did **not** add a guard: D5 shrinks this to the rare case (the secret was empty at capture AND absent from the guest). Real fix = either treat a DB password as fail-closed like a data key, or `ALTER USER` to the regenerated value after the volume restore. 18 DB/root-password fields are in scope; MariaDB fails loudly instead (Access denied), which is the safer half | CC |
|
||||
| **R-200** | Backup & restore | P3 | **The DR password-injection seam has a handler, a route and tests — and no form.** `POST /backup/offbox/inject-password` is routed (`controller/internal/web/server.go:510`) to `offboxInjectPasswordHandler` (`offbox_handlers.go:174-196`) → `InjectOffboxPassword` (`backup/offbox.go:541`). **No template in the repository contains that path or any form posting to it** (grep over `internal/web/templates/`: one unrelated hit, an XSS comment) | **NARROWED** — **PLUMBING COMPLETE** (controller **v0.196.0**); **the FORM is not built — still open** | — | **The tenth instance of this project's built-but-never-wired class, and the exact shape `CLAUDE.md` and `felhom.eu/CLAUDE.md`'s seam-wiring rule were written for:** handler tests that POST directly (`offbox_escrow_test.go:167,180`) prove nothing about reachability. To use the only implemented recovery seam today, a person must hand-craft an authenticated POST with a session cookie and CSRF token. **Note the layering while fixing it:** this form takes a **64-hex repo password**, not a recovery code (`offboxRepoPwPattern`, `offbox.go:543`) — they are different secrets at different layers, and the operator's 2026-08-04 ruling asks for a form that takes **R**. Build the R form and treat this one as the operator/DR fallback it was written as, but **ship it with a render test per branch of whatever gate it sits behind**. Source: `audits/RECON-offsite-dr-chain-2026-08-04.md` §3 link 9 **THE DIAGNOSTIC HALF IS DONE AND IT ANSWERED THE QUESTION.** `--recover-offsite-check` is a `docker exec` escape hatch in the shape of `--print-reset-code`: R on **STDIN** (never argv, never `ps`, never shell history, never a transcript), fetch+unseal via the agent, and a verdict of **two sha256 hashes**. **It compares and never installs** — the recovered password is not written to `offbox/repo_password`; a test asserts the data dir is byte-unchanged and its red-proof (adding the install call) fails it. Confirmed live: `repo_password` mtime still `2026-08-03 07:18:02` after the successful check at `2026-08-04 11:49`. **Exit codes are load-bearing** — `0` match, `2` a clean MISMATCH, `1` a step failed; "it failed" and "it worked and disagreed" must never share a status because only one is a finding about the system. A box with no local password reports distinctly (the rebuilt-box shape, where the next step is to INSTALL rather than compare). **WHAT IS NOT BUILT, deliberately:** no card, no form, no preview, no customer-facing text — building an interface on top of a chain nobody had walked is how the preceding three weeks went wrong. **What remains for this row:** link 9 (the recovered password placed so `WriteOffboxSecrets` keeps it) and the customer-facing shape the operator ruled on 2026-08-04 (yell → R form → preview → proceed), which is now priced against a chain that exists rather than one that is assumed **PART 0 SHIPPED 2026-08-04 (v0.196.0):** `--recover-offsite-install` is the sibling of the check — same fetch/unseal path, same STDIN discipline for R — and it **places** the recovered password via `InjectOffboxPassword`. The confirmation is a SECOND invocation (`--confirm-install`): without it, both hashes print and nothing is written, so the operator sees the comparison before a write is possible. Three outcomes named distinctly: **installed** (no local password — the rebuilt-box shape), **unchanged** (identical key, nothing written), **refused** (a DIFFERENT key present — installing would clobber the key the current repository is encrypted under; exit 2, no force offered). It re-reads the file after writing rather than trusting the call. **Red-proof observed:** removing the confirmation gate makes the dry run write the password. The R-persistence test carries a **positive control** (a planted copy found, then removed and not found). **NOT YET EXERCISED AGAINST A LIVE RECOVERY** — the R-201 drill halted before step 9, so this is unit-proven only. **What remains for this row:** the customer-facing shape the operator ruled on 2026-08-04 (yell → recovery-code form → preview → proceed) | CC |
|
||||
| **R-231** | Backup & restore | P3 | **`/opt/backup/scripts/` on DooPlex is unversioned host state** — found 2026-08-06 while adding the auto-memory store to the backup set. No repository tracks the scripts that protect the recovery chain, so the edit made that day (`CLAUDE_MEMORY_DIR` in `backup-config.sh`, multi-path restic call in `backup-data.sh`) exists only on the box. This is the same class the part-2 session was closing, found inside the fix for it; the change is transcribed in `felhom.eu/workspace/README.md` so it is at least *recorded*. **Two related facts, both understating current safety:** the backup destination (`/mnt/5_hdd/backup`) is on the **same physical disk** as the workspace it protects, and the DooPlex backup set has **no off-site leg** (`sync-hetzner-backups.sh` is jarrs.eu and pulls *from* Hetzner *to* DooPlex). Bringing a root-owned production backup script under version control, and deciding what installs it, is its own scoped change. | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-240** | Backup & restore | P3 | **A backup that covered nothing calls itself „Sikeres".** On a configured box with no app selected for off-site backup, a run reports status `ok` with the warning „Sikeres — nincs mentésre jelölt alkalmazás" — *successful* immediately beside *nothing is selected*. Measured as T4 on the final walk, 2026-08-07; flagged once before (2026-08-06) and deliberately not touched then, because the task that noticed it forbade changing that path. **It is the same rhetorical shape the project has spent a fortnight removing** — R-203's *a warning beside a success is read as a success*, R-234's *„✓ Rendben" over an app that was skipped*, R-225's *unknown rendered as zero* — one notch weaker each time, and this is the weakest and last of them. The state itself is honest and must stay `ok`: an unconfigured box reporting `incomplete` forever is its own defect, pinned by a test. **The defect is the word „Sikeres", not the verdict.** Wording such as „Nincs mentésre jelölt alkalmazás — ez a futás semmit nem mentett" says the same thing without congratulating the customer on it. | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-251** | Backup & restore | P3 | **The recovery listing renders one row per restic TAG, so the customer is shown an "app" they never installed and their data counted twice.** Measured on the fifth walk, 2026-08-07, on the screen the customer reaches after entering R. The snapshot carries tags `felhom-offbox,calibre-web`; the listing renders **two rows** — `calibre-web · 2026-08-07 14:57 · 12.8 MB` and `felhom-offbox · 2026-08-07 14:57 · 12.8 MB`. `felhom-offbox` is the tier's own marker tag, not an application. **The screen's whole job is to let the customer check that what is in the store is what they expect** (*"Nézd át, hogy tényleg azt találod-e itt, amire számítasz"*), and it shows them a stranger's name beside their own data and a total that is double the truth. **Cosmetic, not a data defect** — the restore page correctly offers only `calibre-web`. **Fix:** filter the marker tag out of the listing, or key the rows on the app tag. | **READY** — owner Viktor | — | — | operator |
|
||||
@@ -312,16 +302,14 @@ stopping line that lies.
|
||||
| **R-779** | Security & access | P4 | **[P3-LOW] Part A's "two outside addresses seen as two" is proven through the simulated tunnel only; on the REAL tunnel the second outside address (ep0, one request allowed) was refused by Cloudflare's edge with 403 and never reached the box.** Measured 2026-10-01 19:51 UTC (`audits/visitors-2026-10-01/A/L2-demo-hp-real-tunnel.txt`): no log line on demo-hp; demo-hp's box has no geo restriction in its settings, so a Cloudflare ZONE rule (country or bot, not read) refused a German datacenter address. DooPlex's own address on the real tunnel was seen as itself. **Needs:** one sign-in from a second Hungarian address (the operator's phone off wifi) while DooPlex is locked out — 2 minutes; and say which Cloudflare rule refused ep0. | **WAITING-ON-OPERATOR — rank P3-LOW; owner: operator (a phone), CC reads the logs** **Re-ranked 2026-10-03: P3→P4: a proof gap on the real tunnel; operator-only follow-up.** | — | — | CC + operator |
|
||||
| **R-879** | Security & access | P3 | **A copy of hub.db still holds secrets readable without any key:** each box's hub API key (`hosts.api_key`), each household's owner passphrase and API key (`customer_configs.retrieval_password`, `api_key`), and the PBS-DR token values (`host_pbs_secrets.value`, kept after use). Found 2026-10-05 while answering "what does a hub database backup now contain" (R-133 sealed the console passwords; the off-site passwords were sealed by R-821). A stolen database copy lets an attacker report as any box and read every owner passphrase. `05` §16.2 | READY | — | Seal or hash each with the same seal (the box keys and the passphrases are compared, so a hash may fit; the PBS token is served once, so it could be deleted after use) | CC |
|
||||
|
||||
## Box system & updates — 19 rows (P2 1, P3 15, P4 3)
|
||||
## Box system & updates — 15 rows (P2 1, P3 13, P4 1)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
| **R-812** | Box system & updates | P2 | **[P2] A box never receives operating-system security updates — not the Proxmox host, not the guest's Debian, not its Docker engine.** SEARCHED 2026-10-03 (read-only): `felhom-controller`, `felhom-agent`, `app-catalog-felhom.eu` and `felhom.eu` hold no `apt-get upgrade`, `apt full-upgrade`, `unattended-upgrades`, `pveupgrade` or `needrestart` that runs on a box. The installer aligns the host's Proxmox repositories to no-subscription *"so the box can pull security updates"* and then says plainly *"No upgrades are run"* (`scripts/felhom-host-install.sh:2133-2136`). The guest's Docker engine is installed when the golden is BAKED (`felhom-agent/configs/build-golden.sh:124-125`), so a fresh install gets that week's engine and an installed box keeps it forever. The only `apt full-upgrade` in the project is a by-hand step for the off-site endpoint ep0 (`documentation/runbooks/offsite-endpoint.md:41`), not a box. App images ARE updated (the update arc); the layer under them is not. The intention, with its scope, is **R-808** in `ROADMAP.md`. | **NARROWED 2026-10-04 (evening) — the guest's DOCKER engine slow lane is BUILT and proven live (agent v0.142.0, hub v0.132.0; `11` §5.8): live-restore on everywhere, ring 0 steps under a root-owned mark, ring 1 and undo only by a signed job the wrapper re-verifies; the operator approves each engine set on the System page. LEFT: the kernel lane (R-836); existing boxes (R-840).** Earlier: NARROWED 2026-10-04 (afternoon) — the HOST's Debian fast lane is BUILT and proven live too (agent v0.141.1, hub v0.131.1; `11` §8.2, `audits/os-host-lane-2026-10-04/`): appliances only, never kernel/boot/firmware, after a healthy guest step; fleet view and four alarms (§8.3). LEFT: the Docker and kernel slow lanes (R-836); existing boxes (R-840).** Earlier: NARROWED AGAIN 2026-10-04 (day) — the GUEST's Debian fast lane is BUILT and proven live (agent v0.140.0, hub v0.130.0, installer 1.29.0; `11` §8.1, `audits/os-guest-lane-2026-10-04/`). LEFT: the host, Docker and kernel lanes; the undo (R-842); existing boxes (R-840).** Earlier: NARROWED 2026-10-04 — the SPIKE is done (`11` §7.1, corrections C1–C12, `audits/os-updates-spike-2026-10-04/`); no product code yet. LEFT: the build steps of `11` §8, each with the operator's go; the §5.3 snapshot question is in STATUS; preconditions R-835, R-836, R-837.** | — | — | CC + operator |
|
||||
| **R-862** | Box system & updates | P3 | **Tester 2 cannot take the config bundle until one by-hand bootstrap is done: its `felhom-os-apply` (agent 0.142.0) predates the bundle mode, and no signed job can write a root file on it.** FOUND 2026-10-04 (R-840 build, Part C): the route reaches every box installed from installer 1.31.0 on, and the demo boxes (bootstrapped by CC); Tester 2 has every root file of agent 0.142.0 (installer 1.30.0) and lacks only the R-858 wrapper fix, which matters only for a Docker step it gets solely from a signed job. CC has no route to Tester 2 (its door admits only the operator's WireGuard peer; `felhom-op` cannot become root). The steps: `runbooks/config-bundle.md` "Tester 2". Then CC signs the bundle and reads it back. | **WAITING-ON-OPERATOR** — the operator said (2026-10-04 ~19:05) he will try through his tunnel | the operator's WireGuard tunnel | the operator runs the three bootstrap commands; CC sends the bundle | operator |
|
||||
| **R-35** | Box system & updates | P3 | **Config-apply should not end the customer's session.** The offsite config push bumped `config_version` 10→11 at 16:54:58 and the controller self-restarted (container `StartedAt` 16:54:59Z, back up 16:55:02); in-memory sessions died with it and **customer zero was force-logged-out mid-flow**. **MIGRATED FROM `ROADMAP.md` 2026-08-22 (R-369) — originally filed 2026-07-21, size S, roadmap state `idea`.** Moved verbatim; nothing added or reinterpreted. The roadmap keeps its copy as history, marked moved. | **OPEN — migrated from ROADMAP 2026-08-22, rank unchanged** | — | Direction: **hot-apply the offbox target** (no restart for a config the running process can adopt), or **persist sessions** across restart. The restart itself is by design — the collateral is not. Evidence `controller-log-full.txt` | CC |
|
||||
| **R-50b** | Box system & updates | P3 | **[P2] A root-owned privileged host artifact is delivered unversioned from `main` — "which wrapper is on this host?" is unanswerable.** `configs/felhom-pbs-apply` installs to `/usr/local/sbin/felhom-pbs-apply` (0755 root:root) and is the pinned sudoers vector for `create\ |reconcile\|grant` against `/etc/pve/priv/storage`. It is fetched by `felhom-host-install.sh:1914` via `fetch_raw`, which hits `raw/branch/main/<path>` — **no tag, no pin, no checksum, and no record in the Day-0 artifact manifest**, unlike the agent binary (sha256-vouched) and the golden image. Three consequences: (1) two hosts installed a week apart can carry different privileged wrapper code while both reporting the same agent version; (2) a host hotfixed in place (felhom-pve, 2026-07-18) is indistinguishable from one that fetched the same content — the fleet has no inventory of it; (3) an accidental push to `main` reaches the next install of every host with no review gate between commit and root-owned deployment. | **NARROWED** — **(a) SHIPPED 2026-07-21; (b)/(c) open** — moved from `ROADMAP.md` 2026-10-03: it states a checkable fact about the shipped product, so it is a FINDING (the sorting rule). **Re-ranked 2026-10-03: [P2] → P3 — operator-only; leg (a) shipped.** | — | **Surfaced 2026-07-21 while stopping the R-39 v0.90.1 publish** (`felhom-controller/REPORT.md` §5): the publish was cancelled precisely because the version number would have claimed to carry a fix that in fact rides this unversioned channel. Candidate shapes, in increasing cost: (a) record the wrapper's sha256 in the Day-0 artifact manifest beside the agent binary and have the agent report the installed file's hash, so drift is at least *visible*; (b) `fetch_raw` takes a pinned ref (tag or commit) supplied by the manifest rather than `main`; (c) the wrapper becomes a published generic-registry artifact with the same gate ladder as the agent binary. **(a) is the cheap honest first step and would have caught this class already.** Pairs with R-39 (whose remaining fleet half is specced separately) **(a) SHIPPED 2026-07-21 — hub v0.68.0 + agent v0.91.2.** `ArtifactManifest.WrapperSHA256` + an operator field; agents report the installed wrapper's sha256 each cycle and the host page surfaces a mismatch. **An unknown on EITHER side reads as quiet, never as drift** — lighting every host amber on rollout day is how a warning becomes background noise. Live confirmation of exactly the problem: felhom-pve's July-18 in-place hotfix hashed `2888f2ea…`, matching **no commit anyone could name**; it now reports `104db0a4…` against a vouchable manifest value. **(b)/(c) REMAIN OPEN:** the wrapper is still fetched unversioned from `raw/branch/main` — this makes drift *visible*, it does not fix the channel. Also recorded: the 0440 sudoers file is not agent-readable, so its drift stays invisible. **Flips (2026-10-03):** `00` §A "The installer is PUBLISHED, not pushed" — the same discipline for the privileged wrappers. **Re-ranked 2026-10-03:** [P2] → P3: operator-only; (a) shipped 2026-07-21 (the report carries the wrapper sha256), (b)/(c) open. **Finding-shaped** — an R-424 instance; check against today's product before building. | CC |
|
||||
| **R-78** | Box system & updates | P3 | **`local_api` authority ruling — auto-reconcile vs detect-only** **MIGRATED FROM `ROADMAP.md` 2026-08-22 (R-369) — originally filed 2026-07-25, size M, roadmap state `idea (deferred OUT of R-77 on purpose)`.** Moved verbatim; nothing added or reinterpreted. The roadmap keeps its copy as history, marked moved. **An OWED OPERATOR DECISION, not a defect — moved because an owed ruling hidden among feature ideas is the shape this session exists to remove.** | **OPEN — migrated from ROADMAP 2026-08-22, rank unchanged** | — | R-77 ships detection because the fix is genuinely undecided, and **both directions can lose customer-visible function**. **Direction 1 (today):** `controller.yaml` wins and drift is silent → the 2026-07-25 island migration blinded the whole fleet's control plane for 17.5 h (drive gate, guest-reboot recovery, quiesce/backup all degrade). R-77 makes that loud but does not stop it recurring. **Direction 2 (`bootstrap.json` wins, auto-reconcile on boot):** a guest whose `controller.yaml` is CORRECT and whose `bootstrap.json` is stale — a half-completed re-provision, a hand-repaired guest, a setup-wizard box — gets a **working channel clobbered on the next restart**, fleet-wide and silently, during a routine deploy. That is not obviously better than the bug. Needs a spike: which writer is authoritative per field (endpoint vs fingerprint vs token — `mergeLocalAPI` replaces the whole block, so they cannot be reconciled independently today), whether the agent side should stamp a generation/mtime so 'newer wins' is even expressible, and whether reconcile should require an operator ack. Until then the drift alert plus a manual edit is the supported path. | CC |
|
||||
| **R-121** | Box system & updates | P3 | **A BOX's installed agent can sit releases behind the vouched one and nothing notices — the R-120 gate does not cover it.** demo-hp ran agent **0.113.0** while the hub vouched **0.116.0**, through the whole R-116/R-117 arc, and no signal existed on any channel | **READY (S) — NEW 2026-07-30** | — | **Fourth instance of the drift family** (R-111 golden's agent 17 releases behind, R-115 built+deployed but never published, R-120 golden a controller behind — and now installed-vs-vouched on a live box). **Confirmed at source that R-120's gate cannot catch it:** `hub/internal/web/configs.go:1165-1169` compares `goldenVer` against `store.NewestReportedControllerVersion()` — it is a **golden-artifact vs fleet-CONTROLLER** check and says nothing about the agent installed on a box. **`MinAgent` does not cover it either:** it is used to HOLD the controller floor for a box whose agent is too old (`hub/internal/api/handler.go:530-538`, `store.go:1857`) — protective, not an alarm — and demo-hp's 0.113.0 **equalled** `min_agent` 0.113.0, so even a floor comparison was satisfied. **The cost, measured:** R-117's whole subject is the R-113 conjunction, which landed in **0.114.0** — so the designated drill host could not exercise the code under investigation at all, and the R-117 spike had to route every predicate result through an out-of-repo probe built from `main` instead of the installed agent (`audits/SPIKE-r117-bind-liveness-2026-07-30.md` §1, §2.3). Discovered because the R-117 task made bringing the box current an explicit prerequisite. **Fix shape (not implemented):** the hub already receives `AgentVersion` on every host report, and already has semver comparison in Go — the missing piece is a checker comparing reported agent vs the **vouched** agent and surfacing it, operator-tier. Note the honest tension: a box legitimately lags between publish and deploy, so this wants a staleness window rather than an instant alarm | CC |
|
||||
| **R-190** | Box system & updates | P3 | **A storage ACL that demonstrably WORKED in the morning was gone by mid-morning, and nothing recorded its removal.** On demo-felhom, a `vzdump` by `felhom-agent@pve!agent` with `--storage felhom-backup` completed **OK at 04:44:50 CEST 2026-08-03** (task log read in full). From **09:24:56** the same path returned `HTTP 403 … missing privilege Datastore.Allocate at /storage/felhom-backup`, six times through the day, until the grant was re-applied by hand at 18:54. By ~14:50 `pveum acl list` showed **no row at all** for that path | **NARROWED** — **MITIGATION SHIPPED 2026-08-04** (agent **v0.124.0 → v0.124.1**) — **MECHANISM STILL OPEN** | — | **Why this is not just R-185 restated:** R-185's mechanism (the installer's Scenario-F arm resolves a pre-existing target without granting) explains a box that NEVER had the grant. This box HAD it and lost it, inside five hours, with the machine up throughout. **Ruled out, each by measurement:** a host reinstall (`uptime` = 12 days); any `pveum`/ACL/`user.cfg` activity in syslog between 04:00 and 10:00 (none); any ACL entry in `/cluster/log` (none). **Correlated, not established:** `host_leaf_changed` at 09:15 and `controller_started` at 09:19 — guest 9201 was reprovisioned nine minutes before the first 403. PVE removes ACLs at `/vms/<vmid>` when a guest is destroyed (`AccessControl::remove_vm_access`, the F-LEAK mechanism); whether any path can take a `/storage/<id>` row with it has NOT been established and is the first thing to check. **Why it matters more than the grant did:** a permission that can vanish silently makes every ACL-based guarantee on these hosts provisional, and the agent's new store-grant probe (v0.123.0) now detects the STATE but says nothing about the TRANSITION. **Worth pairing with:** whether the probe should report a grant it once had and no longer has as a distinct, louder signal than one it never had **THE ROW NOW REFLECTS THE MITIGATION, NOT THE CAUSE — stated plainly because the two are different things.** The box is resilient; the loss is still unexplained. **Mitigation:** when the store-grant probe finds the grant absent, the agent runs the EXISTING root wrapper (`felhom-backup-target-apply grant <id>`) and **re-reads once** to confirm — the pbsdr R-22 self-grant shape, including its restraint. **No new privileged surface:** the sudoers vector `grant *` already covers any storage id (confirmed in `configs/felhom-agent.sudoers`, not assumed), and the verb already grants BOTH user and token. The verb existed, was permitted, and had only ever been called at storage CREATION — the *built but never wired* shape in a verb rather than a seam, this project's seventh instance. Bounded at one attempt per tier per hour (a storage can be unreadable for reasons an ACL cannot fix; re-granting every cycle is a repair loop wearing a fix's clothes). **THE RECORD IS THE HALF THIS ROW IS ABOUT, and v0.124.0 got it wrong in production while every unit test passed.** It reported degraded for *one cycle* — meaning the probe call that repaired. But `probeAll` is invoked INDEPENDENTLY by the self-check log and by the collector building a host-report: on the box the repairing call was the log's (`09:39:34`, journal shows the repair and `degraded=1`) and the report three seconds later found the grant present and sent **`ok`**. The agent's journal had the record, the hub had nothing, and the operator would have learned nothing — the exact silence this row exists for, re-created inside its own mitigation. **v0.124.1** replaces it with a latch on TIME (20 min > the 900 s report interval), so at least one report must carry it. **PROVEN LIVE, twice, on demo-felhom** (grant deleted by hand, both rows): agent logs `store-grant: GRANT WAS MISSING AND HAS BEEN SELF-REPAIRED — investigate the loss (R-190) target=felhom-backup privilege=Datastore.AllocateSpace action="felhom-backup-target-apply grant felhom-backup" confirmed_by=re-read`; the ACL rows return; and on v0.124.1 the **host-report at 08:00:30Z carried `status=degraded`** with the explanation and the hub raised `agent_capability_degraded` and **e-mailed the operator** at 08:00:40. Nothing new was built to carry it — the hub's existing ok→degraded→ok edge is the channel, and the text rides `Feature` because that is the field the hub interpolates into the e-mail (`Reason` does not travel). **PART 3 — the single bounded pass at the mechanism, with the negatives named.** The lead §8.6 nominated is real as a CLASS and is documented in our own installer: *"`pveum user token remove` purges the token's ACL, so re-applying post-rotate is mandatory"*. **It does NOT fit this box.** A rotation purges ALL of the token's ACLs and mints a NEW secret; demo-felhom's token still authenticates with the same secret (`--selftest` OK), it retained its other three storage grants throughout, and only `felhom-backup` was refused. No installer run is evidenced (no 2026Line truncated
|
||||
| **R-242** | Box system & updates | P3 | **A controller release that changes customer-visible behaviour is not delivered until a golden carries it — and nothing enforces that.** R-239 is the symptom; this is the mechanism, recorded **2026-08-07** and deliberately **NOT built** (the task that found it scoped it as a record-only item). **Two releases went out without a golden and the gap was invisible until a walk measured it from the customer's side**: v0.204.0 (R-237) and v0.205.0 (R-234) were written, tested, pushed, and CHANGELOG'd, and every one of those steps passed while a machine installed that night received neither. The register said CLOSED; the fleet said otherwise. **Nothing in the release path knows a golden exists.** The version bump, the image push, the CHANGELOG entry and the register closure are all repo-local; the manifest's `golden_version` is edited by a separate operator act, in a different repo, with no link back. **Proposed shapes, cheapest first — the choice is the operator's and is not taken here.** (a) **A release-path checklist step** — one line in the controller's end-of-session checklist: *a release that changes customer-visible behaviour is not finished until a golden carries it or a register row says why not.* Costs nothing, catches nothing mechanically. (b) **A gate in `repo_gates.py`** comparing the manifest's `golden_version` against the newest released controller and FAILING (or warning) past a tolerance of one minor. Mechanical, runs on every push, and would have fired the morning after v0.204.0. (c) **A hub-side checker** — the hub already knows every box's running controller version from `/hosts` and the vouched golden from the manifest; a periodic comparison against the newest published image would catch drift the repo cannot see, including a vouch that was made and then rolled back. **Earliest catch: (b).** It fires on the push that creates the gap, before any box is installed, and it needs no live fleet. **(c) catches strictly more but only after boxes exist.** (a) is worth doing regardless because it is free. **Not built. No gate was written this session.** **⚠ IT RECURRED WITHIN A DAY, WHICH IS THE ARGUMENT FOR BUILDING IT.** Controller **v0.206.0** shipped the R-241 fixes on 2026-08-07 while the vouched golden still carried **0.205.0** — so a machine installed on the morning of 2026-08-08 would have received neither. Third occurrence of the shape in three days (R-111/R-115/R-120 are the older family). **✅ SHAPE (b) BUILT 2026-08-08 — `scripts/golden_currency_gate.py`, registered in `repo_gates.py` as gate 7.** **It was shown FAILING against that exact state before anything was baked**, which is its red-proof and the reason its own introducing push needed `--no-verify` (stated in the session report rather than worked around): `newest released controller : 0.206.0 / newest golden baked : 0.205.0 → CONVICTED`. **IT IS `--fast`, AND THAT FORCED ITS DESIGN:** both `.githooks/pre-push` AND CI run `repo_gates.py --fast`, so a non-fast gate would run in NEITHER — the R-29 census failure this runner exists to end. **THEREFORE IT CHECKS THE BAKE, NOT THE VOUCH**, because the vouched version lives only in the hub's `hub_settings` with no copy in git, and putting a copy there would create a second source of truth that can drift — a green gate over a false claim being the worst outcome available. **A bake without a vouch still passes: that half is NOT closed and stays on this row.** It also compares versions rather than behaviour, so a release changing nothing customer-visible trips it too — accepted deliberately, because judging that by hand is what failed three times and the cost of a false trip is one bake; a waiver belongs here, never in a habit of bypassing. **VOUCHED 2026-08-08 with the operator's approval** — golden `0.206.0` / sha `c85230b4…108e`; `agent_version` and `min_agent` both stayed `0.127.0`, and `wrapper_sha256` was carried through explicitly because the handler clears it when omitted. **The gate was CONVICTED before the bake and OK after it** — red→green on the same command, which is its proof that it measures something real. **⚠ THE GATE FIRED FOR REAL, 2026-08-08 — and it was right.** Controller **v0.207.0** (R-249/R-252/R-253) is released, tested and pushed, and **no golden carries it** — the newest bake is 0.206.0 — so `golden_currency_gate.py` FAILED, saying exactly the true thing: *a machine installed right now would receive v0.206.0*. **The `felhom.eu` push therefore used `git push --no-verify`, declared here, in the commit message and in the session report.** A bypass and NOT a waiver, deliberately: the gate offers a waiver only for a release that *deliberately needs no golden*, and this one needs one. **Owed: bake golden 0.207.0 and vouch it** (`RUNBOOK-manual-build.md` §4.1; the vouch is a three-field change). **This row's own remaining half is unchanged — nothing gates the VOUCH itself.** **✅ THE OWEDLine truncated
|
||||
| **R-274** | Box system & updates | P3 | **A local golden is adopted with NO version and NO checksum check, so a reinstall can silently come up releases behind.** `felhom-host-install.sh` step 7: `if [[ -n "$GOLDEN_VOLID" ]] && ! $FORCE_GITEA_GOLDEN; then log_skip "using local golden"; return 0; fi` — **the hub manifest's `golden.sha256`, whose entire purpose is to vouch from a different trust root than Gitea, is consulted only on the FETCH path.** A locally-present archive bypasses the vouch: no version compare, no digest, no warning. On demo-hp 2026-08-09 the preflight selected `local:backup/vzdump-lxc-9100-2026_08_03-07_33_00.tar.zst`, whose baked marker reads `felhom-controller:`**`0.192.0`**, against a vouched golden of **0.210.0** — 18 releases stale. **The sharp consequence:** 0.192.0 is **below 0.200.0, where R-193's off-site recovery SCREEN shipped**, so a customer reinstalled today returns on a controller that cannot run the recovery ceremony their data depends on; it is also born below the managed-update floor (0.200.0), and the updater's auto-target is the floor, never the newest. **It compounds with the teardown**, which deliberately keeps the old golden (*"golden vzdump left in place"*). This is the R-111/R-115/R-120 drift family one layer down: the R-120 gate guards what may be VOUCHED, nothing guards what an install TAKES. **OBSERVED 2026-08-09, AND THE RESULT NARROWS THE ROW — recorded because it partly refutes what was written above.** On the RESUME path step 7 **fetched the vouched 0.210.0 correctly** (`fetching golden v0.210.0 from Gitea`), because `--resume` skips preflight and preflight is where local auto-discovery sets `GOLDEN_VOLID` (the GL6-F4 comment says so). **So the fresh-install and resume paths disagree on golden selection, and the resume path is the safe one.** Discovery is `… | sort | tail -1`, i.e. the NEWEST local archive by filename — a sensible heuristic, **and still no comparison against the manifest's version or sha**. The defect therefore stands as: *a box whose newest local golden predates the vouched one installs stale, silently* — which is exactly the state demo-hp was in before this run (newest local 0.192.0 vs vouched 0.210.0). It is now masked on this box because the freshly fetched 0.210.0 is the newest — **correct by recency, not by verification**. There are now **three** goldens on `local` (07-21, 08-03, 08-09), because the teardown keeps them. **Still not observed: a FRESH (non-resume) install taking a stale local golden.** | **VERIFY** (2026-10-03 triage: Local golden is now checked against the hub manifest's version and sha before use (GOLDEN_CHECK_WHY, R-297) — felhom.eu/scripts/felhom-host-install.sh:2855-2905) — **READY (S) — NEW 2026-08-09, NARROWED same day** | — | Compare the local golden's version/sha against the manifest and refuse or re-fetch on mismatch; say so in the BYO disclosure, which today lists only what the install CREATES, never what it REUSES | CC |
|
||||
@@ -329,14 +317,12 @@ stopping line that lies.
|
||||
| **R-444** | Box system & updates | P3 | **[P3-LOW] Nothing runs `pct fstrim` on the fleet, and demo-hp's thin pool was carrying ~23.8 GB of blocks the guest had already freed.** MEASURED 2026-09-01 during this spike's teardown: the run itself added ~1.05 GiB that `local-lvm` did not reclaim on delete (68.97% -> 70.91%); `fstrim` INSIDE the unprivileged container is refused (`FITRIM ioctl failed: Operation not permitted`, all three mounts); `pct fstrim 9201` from the PVE host then trimmed **30.2 GiB + 57 GiB** and took `local-lvm` to **26.78%** — **23.8 GB BELOW this run's own starting point**, i.e. the surplus was long-standing, not ours. **Why it is not merely housekeeping:** a thin pool that only ever grows can reach 100% from DELETED data alone, and a full thin pool takes every guest on the host read-only. demo-hp had 16.4 GB free before the trim. **Not urgent, and the row says so** — but the appliance has no periodic trim and no operator surface reports the gap between guest-free and pool-used. Owner: **CC.** `audits/SPIKE-app-update-2026-09-01.md` | **OPEN — rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-468** | Box system & updates | P3 | **[P3-LOW] THE GOLDEN WAIVER — goldens on a cadence, not per release (operator ruling 2026-09-13).** 25 goldens in 26 days in August, almost one per release, because `golden_currency_gate.py` trips on every release by design and the only honest ways past it were a bake or a declared `--no-verify` (thirteen by 2026-09-01, R-404/R-417). **The ruling: bake WEEKLY, and always before any drill or fresh install.** Every release still raises the FLOOR, so both demo boxes keep getting each release in ~20 s; only the golden — which protects a fresh install and nothing else — moves to a cadence. **The mechanism (built 2026-09-13):** `documentation/tests/golden-waiver. **⚠ CORRECTED THE SAME DAY (R-472): between bakes the floor does NOT carry a release — the hub holds any floor above the vouched golden (publish-train rule 1), so releases between bakes reach the demo boxes only by hand-deploy.**yml`, four lines (`issued`, `expires`, `reason`, `register_row: R-468`), read by the gate. While valid, a golden BEHIND the record makes the gate print a loud ADVISORY and exit 0; when it expires the gate is red again until someone bakes or renews. **The 14-day cap is enforced by the gate, not the runbook** — a longer, undated, unparseable, reason-less or row-less waiver is INCONCLUSIVE (exit 2), never 0 and never silently ignored. **It never covers a golden that is UNRECORDED (R-385)** — that is not a cadence choice. **A dated waiver cannot be forgotten; it just expires** — the difference from R-242's original rule, which recurred the day after it was written. Tests: `scripts/test_golden_currency_gate.py` cases 5–15 (E/F/G/H, a 15-day, absent, unparseable, bad-row and empty-reason waiver each 2; the R-421 decoy — a file saying only `expires` — 2). **This is a PRE-CUSTOMER arrangement: the first external install retires it** (delete the file in that commit). Cadence written into `RUNBOOK-manual-build.md` §4.2 and the `felhom.eu` end-of-session checklist. **Does NOT touch R-242's open half (nothing gates the VOUCH).** | **WATCHING — rank P3-LOW; owner: CC (renew ≤ 14 days or bake); retire at the first external install** | — | — | CC |
|
||||
| **R-531** | Box system & updates | P3 | **[P3-LOW] Three supervisor facts measured live and not pinned: restart timing during a deploy was not measured; restarts before the hub first sees the stanza produce no `controller_restarted_by_agent`; deliberate operator kills spend the crash-loop budget.** MEASURED 2026-09-15 on 9201: after 3 test restarts in 13 minutes the 4th kill tripped the 30-minute pause and the dashboard stayed down (the guard as designed, `A4-kill-middeploy-9201.txt`). The hub checker seeds silently on first sight, so the three restarts before the first v0.131.0 report emitted nothing (only the crash-loop did). **What it needs:** a deploy-kill timing on a fresh budget; the operator's view whether a restart after minutes of uptime should count toward the budget **MEASURED 2026-09-16 on the drill box, both halves.** (1) **Timing during a deploy (F9'):** the controller was killed 5 s into a deploy on an EMPTY budget; the agent saw it on the next sweep, confirmed on the one after, and the dashboard answered 200 again **37 s** after the kill; the interrupted app ended `not_deployed`, not stuck. (2) **The budget's shape (F9''):** three further kills at idle, 20 minutes apart, recovered in **61 s / 41 s / 61 s** - and NONE of them accumulated, because the window is 15 minutes. Four restarts this session, zero pauses, zero crash-loop events. **So the brake catches a FAST loop and is blind to a SLOW one:** a controller dying every 20 minutes is restarted forever and the only trace is an `info` event that mails nobody. That is a design question for the operator (leave it / add a longer second counter / raise the severity of the Nth restart in a day), and this session deliberately measured it without changing it. Evidence: `audits/evidence-drill-0243-2026-09-16/phase2-f9prime.txt` and `phase2-f9dprime.txt`. | **READY — rank P3-LOW; owner: CC (measure) · operator (budget rule)** | — | — | CC + operator |
|
||||
| **R-184** | Box system & updates | P4 | **Nothing prevents the hub from vouching an agent version that was never released.** The R-115 gate proves every RELEASED version is installable, but it works from git tags — so a hub artifact-manifest entry naming a version with no tag and no package is invisible to it. The installer would then die at step 5 on a virgin machine, as root | **READY (S) — NEW 2026-08-03** | — | **Filed BECAUSE the R-115 gate deliberately does not cover it, rather than leaving the gap unstated.** CI cannot check it: the hub's `/api/v1/artifacts/<customer>` answers **401** without a per-customer retrieval passphrase and Gitea's package **listing** api answers **401** without a token (both measured 2026-08-03, P-C), so a credential-free gate can ask *"is this version installable"* but never *"which version is vouched"*. **Two shapes, and the second is better:** (a) give CI a hub credential — expands what CI can reach, and is the operator's call not a gate author's; (b) **validate at vouch time, in the hub**: the operator UI's Day-0 artifact form refuses a version whose package is not downloadable. (b) fails closed at the moment of the decision, needs no new credential anywhere, and puts the check where the mistake is actually made. **Exposure is low and should be said so:** vouching is a deliberate operator action against a version they have just released, and R-115's release path now makes released-but-unpublished nearly impossible. This is the residue, not the main risk | CC |
|
||||
| **R-194** | Box system & updates | P4 | **PVE's permission cache delays every grant-state verdict by an unknown amount, so "the agent can read it" and "the ACL exists" are not the same measurement.** Observed twice while validating R-190's self-repair on demo-felhom 2026-08-04: both ACL rows for `/storage/felhom-backup` were deleted, and `GET /access/permissions` continued to report `Datastore.AllocateSpace` present — for **~40 s** in one run and **~16 minutes** in another. During that window the capability probe reads healthy and the self-repair does not fire | **OPEN** | — | **Why it matters beyond the delay:** it puts a floor under how fast a lost grant can be noticed, it makes any single permission read a lagging indicator, and — the interesting part — **it is a candidate contributor to R-190's own timeline**: a grant removed at an unknown moment could keep working until a cache expiry, which is exactly the shape of *worked at 04:44, refused at 09:24*. That does not explain what removed it, but it may explain when the refusal SURFACED, and the two have been treated as the same instant. **Not a defect in our code** — it is PVE behaviour, and the mitigation already tolerates it (the repair fires on the next probe after the cache clears). **What is worth deciding:** whether the store-grant probe should ALSO consult the storage content listing as a second signal, since that appeared to reflect the loss immediately (`{"data":[]}` while the permission read still said present) — two signals disagreeing is itself information, and today only one of them is read | CC |
|
||||
| **R-373** | Box system & updates | P4 | **`SysDataGrowGB` is the intended lever for the system-data volume, it works, and nothing sets it.** Written down 2026-08-02 in `audits/SPIKE-recovery-unit-space-2026-08-02.md:230-232`, under an explicit *"### Not filed"* heading: the 20 G / 50 G mismatch was ruled a tier-sizing decision rather than a defect, *"`SysDataGrowGB` is the intended lever and it works; nothing sets it."* A lever with no caller is the same shape as R-368's comment — a setting that names a behaviour nothing invokes. **Age when filed: 20 days.** | **OPEN — LOW** | R-368 (same shape) | Either wire it to something an operator can reach, or remove it and record the sizing decision where a reader will meet it. | CC |
|
||||
| **R-836** | Box system & updates | P3 | **A new host kernel that hangs before userspace stays the GRUB default: `--next-boot` is not a one-shot on these hosts.** MEASURED 2026-10-04 on demo-hp (operator's word, 2 reboots): both demo hosts boot UEFI + GRUB without proxmox-boot-tool ESPs; installing a kernel makes it the default at once; `kernel pin <new> --next-boot` writes an ordinary `GRUB_DEFAULT`, and `proxmox-boot-cleanup.service` clears it only after a boot reaches userspace. With the old kernel pinned FIRST, the fallback after a good boot worked (new 60 s, old 76 s). READ FROM THE CODE, not measured: a hang leaves the new kernel default on every power cycle. Only `softdog` runs (useless before userspace); demo-hp's `sp5100_tco` ships unloaded, untested. Fix direction for the slow lane: GRUB's own one-shot (`GRUB_DEFAULT=saved` + `grub-reboot`) with the old kernel saved — to be measured, including Secure Boot (ON on demo-hp). `audits/os-updates-spike-2026-10-04/partH/` | **NARROWED 2026-10-04 (os-host-lane Part E, operator's word before each of 2 reboots) — GRUB's own one-shot is NOT a one-shot here either.** `GRUB_DEFAULT=saved` (old kernel saved) + `grub-reboot <new>`: boot 1 → new kernel, **Secure Boot ON and fine**; but GRUB could not clear `next_entry` (`grub-reboot` itself warns: *environment block on lvm device … will remain the default until manually cleared*; `/boot` is ext4 on LVM `pve-root`), so boot 2 (no command) → **the new kernel again**. `kernel.panic = 0`: a panic leaves the host stopped (R-851). `sp5100_tco` LOADS and answers (`SP5100 TCO timer`, 60 s, inactive, nowayout 0; read from sysfs, never opened, unloaded) — a hardware watchdog exists on demo-hp, but nothing arms it before userspace. demo-hp left on 7.0.14-20 with that as the saved default. **LEFT (fix direction):** a GRUB env block GRUB can write (on the ESP, vfat) or a userspace "boot good" step that rewrites the default, plus arming `sp5100_tco`; to be measured before the kernel slow lane. `audits/os-host-lane-2026-10-04/partE/` **READY — owner: CC + operator (reboots).** | — | — | CC |
|
||||
| **R-853** | Box system & updates | P3 | **After a boot the box's versions and crash facts reach the hub up to ~15 minutes late.** MEASURED 2026-10-04 on demo-hp (crash-guard test): the agent's first report after a boot has no `system.facts` — the facts read needs a RUNNING customer guest (`firstGuest`), the guest starts ~1–2 min after the agent, and the failed read is cached for 10 minutes; so the HOST half (the crash guard, the kernel) is lost too. The crash events arrived 15 min after the boot (17:17 → 17:32 CEST); nothing was lost (the guard keeps 7 days). Fix direction: the facts mode reads the host without a guest (guest fields `unknown`), and a failed read is not cached. `audits/os-docker-crash-2026-10-04/partC/c4-hub-events.txt` | **READY — owner: CC** | — | — | CC |
|
||||
| **R-839** | Box system & updates | P3 | **After a Docker restart that stopped every container, the boot sweep HELD an app whose `app.yaml` `HDD_PATH` names its user folder instead of the drive.** MEASURED 2026-10-04 on scratch 9202 (paperless-ngx): `bootrecon` logged `drive /mnt/felhom-drives/scratch_hdd/userdata/paperless-ngx is not a live mountpoint — NOT starting it`, although the drive `/mnt/felhom-drives/scratch_hdd` IS a mountpoint; the app stayed down until started by hand. The customer boxes' 9201s showed no hold. **Not diagnosed:** which writer put a per-app path in `HDD_PATH` on 9202, and whether a household box can get it. `audits/os-updates-spike-2026-10-04/partG/SUMMARY.md` | **READY — diagnose; owner: CC** | — | — | CC |
|
||||
|
||||
## Monitoring & notifications — 28 rows (P2 3, P3 17, P4 8)
|
||||
## Monitoring & notifications — 26 rows (P2 3, P3 15, P4 8)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
@@ -353,9 +339,7 @@ stopping line that lies.
|
||||
| **R-521** | Monitoring & notifications | P3 | **[P3-LOW] One unplugged drive sends the operator five e-mails and the household none.** MEASURED 2026-09-14 (BIGNIGHT F4, VM 333): `storage_disconnected (error)` at 21:58:02 CEST plus `app_start_failed (warning)` for each of the four apps the drive carries at 21:58:15, each with its own operator mail (hub log: five `Operator email sent`). The customer's mailbox (`tester1@felhom.eu`, read through the connector) received nothing; the household learns of it only on the dashboard, which is honest and says what to do. The apps' stop is a consequence of the drive event, so the four warnings add no information. **Fix shape:** suppress `app_start_failed` for apps stopped by a `storage_disconnected` (the dead-app check already knows the reason — „Hiányzó tárhely"), and decide whether a household gets a mail for a lost drive. **F6, 40 min later, the opposite failure:** a second, separate drive loss (20:38:33Z) produced `storage_disconnected (error)` and four `app_start_failed`, and the hub logged `Operator email suppressed … cooldown` for all five — **no mail at all for the second unplug**; only `health_degraded (warning)` mailed. A per-key cooldown that outlives the recovery (`storage_reconnected` came between them) silences a new incident. **F7:** the system disk at 95 % produced only `health_degraded (warning)`, whose operator mail was **suppressed by the cooldown** left by F6's `health_degraded` 15 minutes earlier; no disk-specific event reached the hub at all — the operator was not told the disk was nearly full. | **READY — rank P3-LOW; owner: CC (controller) · operator (customer mail policy)** | — | — | CC + operator |
|
||||
| **R-522** | Monitoring & notifications | P3 | **[P3-LOW] While the box has no internet, the dashboard's „Cloudflare Tunnel" tile keeps saying „Fut", and no page tells the household the box is offline.** MEASURED 2026-09-14 (BIGNIGHT F8, VM 333): VM 333's traffic off the LAN and to the hub was dropped at demo-hp's bridge 21:08:36 → 21:26:07Z. Throughout, the LAN dashboard (probed every 26 s from demo-hp) answered 200 and, polled every 2 min, showed no banner and the tile „Cloudflare Tunnel — Biztonságos internetkapcsolat — a szerver portnyitás nélkül érhető el kívülről. · **Fut** · Védett"; meanwhile cloudflared logged ≈ 20 errors every 2 minutes, the public name answered 530, and the controller logged `[report] Push failed … context deadline exceeded` and `Job hub-report failed: hub push failed after 3 attempts`. The tile reports the container, not the connection. A household whose remote access is gone sees „Fut". **Fix shape:** the tile reads the tunnel's connection state (cloudflared's registered connections or the report push result) and says „Nincs internetkapcsolat" when either fails. | **READY — rank P3-LOW; owner: CC (controller)** | — | — | CC |
|
||||
| **R-547** | Monitoring & notifications | P3 | **[P3-LOW] A disk that fills and empties between sweeps is never mentioned to anyone: `disk_critical` is defined at ≥95 % used, but the fill-watch runs once a day.** MEASURED 2026-09-17 (chaos night) on a fresh box (`tester-1-022354`, controller 0.245.0): the customer guest’s root filesystem was held at **96 % for ten minutes** (29 G used, 1.5 G free) and **no alarm of any kind fired** — checked twice, once by the round’s own runner and once independently after the fill was released. Cause, established from the ladder BEFORE the round rather than after: `fillwatch` runs **daily at 03:30** plus once ~90 s after a controller start, so a ten-minute window contains no check unless a restart lands inside it. The timing here was almost comic — the controller restarted at 21:28 after the previous round’s power cut, so its one opportunistic check ran about **twenty seconds before** the disk filled. Meanwhile all twelve apps kept serving and the background household loop logged 12 operations with 0 failures, so nothing else would have hinted at it either. **This is the ladder working as designed, not a missed alarm** — the fill-watch is a daily sweep, not a monitor. It is filed because the honest answer to „would the household be told their disk is full?” is **no, unless the controller happens to restart while it is full**, and that answer is not written down anywhere. **Fix shape (one of):** sample the fill more often than daily (a cheap `statfs` on the 5-minute health pass would do it); or say plainly in `08-alarm-ladder.md` that a transient full disk is out of scope. Evidence: `audits/evidence-chaos-night-2026-09-17/round-3.txt`. | **READY — rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-573** | Monitoring & notifications | P3 | **[P3-LOW] The agent-channel and endpoint-drift banners reach the dashboard as finished Hungarian, so they stay Hungarian on an English page.** FOUND 2026-09-18 by localisation slice 2 release A (R-557): `Alert` now carries `MessageKey`+`MessageArgs` and `GetAlerts(lang)` renders them, but `SetAgentChannelAlert(down, msg)` and `SetEndpointDriftAlert(drift, msg)` take a composed SENTENCE from the channel-health checker (`internal/channelhealth`), so those two banners set `Message` directly and render verbatim in both languages. They are the two most operationally important banners there are (R-77, the 17.5 h 2026-07-25 outage). **Fix shape:** the checker passes a KIND plus its parameters instead of a sentence — the same shape R-553 used for the off-site warning — and the two setters build `MessageKey`/`MessageArgs`. One render case per banner in English. | **READY - rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-585** | Monitoring & notifications | P3 | **[P3-LOW] Six event producers still send Hungarian only, so an English household can see one Hungarian line in some mails.** FOUND 2026-09-18 by localisation slice 3 Part B (R-558, controller v0.256.0), which converted 19 of the customer-facing producers and left these: `backup_failed`, `db_dump_failed`, `backup_integrity_ok`, `backup_integrity_failed`, `offbox_enlarge_blocked` and `local_api_endpoint_drift`. **Why they were left:** each receives its sentence already FINISHED from another package, so the key and its arguments no longer exist by the time the notifier sees it — converting them means changing their callers, not the notifier. The 15 operator-tier types are deliberately excluded and are NOT part of this row: the operator reads Hungarian. **Why it matters more than it looks: `offbox_enlarge_blocked` has no `customerMessages` entry on the hub**, so its raw sentence IS the household's mail rather than an extra line under a translated headline — for that one type an English household gets a wholly Hungarian mail, not a mostly-English one. **Fix shape:** push the key and its arguments down from each caller (the shape slice 2 release B already used for errors, `util.MsgError`), then add each to the `convertedProducers` table in `internal/notify/message_customer_test.go`, which is the list both language tests walk. | **READY - rank P3-LOW; owner: CC** | — | — | CC |
|
||||
| **R-635** | Monitoring & notifications | P3 | **[P1-HIGH] `romm 5.3.0` does not fit the memory the template gives it, and the guarded Update called that a success — the app has been OOM-crash-looping on demo-hp for six hours at ~500% CPU.** FOUND 2026-09-22 17:37 **because the operator heard the fans**, which is the only reason it was found at all. `romm` was promoted `5.0.0 -> 5.3.0` on the live catalog that morning (`audits/PROBE-FIX-2026-09-22.md`) after the edge was PROVEN on scratch guest 9202, and the guarded Update was then pressed on demo-hp guest 9201 at **09:08:22Z**, reaching **`done` in 74.8 s** with the app `running`. **It ran clean for two hours.** The first worker kill is at **11:09:20Z**; by 15:38Z there had been **4,530** of them — `Worker (pid:…) was sent SIGKILL! Perhaps out of memory?` — with `docker inspect` reading **`OOMKilled: true`**, the container pinned at **457 MiB of its 512 MiB limit**, and `docker stats` showing **499.51% CPU**. The host's load average sat at **5.2 while otherwise idle**. `rq_cron` is killed and restarted every few seconds in a permanent storm. **THREE THINGS THIS ESTABLISHES, and the third is the one that changes how promotions are judged.** (1) The template's own comment says *`RAM: ~300MB (mem_limit: 1024M total — romm 512M + mariadb 384M + redis 128M)`*; **5.3.0 needs more than 512M and the template was not re-sized when the version moved.** A version move is not only an `image:` line. (2) **The update reported `done` and the app reads `running`**, because nginx answers `GET /` with 200 while the gunicorn workers behind it are being killed — a THIRD variant of the R-618/R-630 theme: the probe is right, the port is right, and the answer is still a false green. (3) **THE ALARM DID FIRE, AND MY FIRST WRITE-UP OF THIS ROW SAID IT DID NOT — CORRECTED 2026-09-22 BY THE OPERATOR, WHO PRODUCED THE MAILS.** The controller HAS an OOM detector (`main.go:821`, `[deadapp] romm: container romm was OOM-killed`, evaluated every 30 s), it emits **`app_oom`** (`notifier.go:726`), the hub allow-lists it (`dispatcher.go:636`) and dispatched it to the OPERATOR channel as **SENT** — `admin@felhom.eu` received **`[Felhom] ⚠️ demo-hp: app_oom`** at **11:09 CEST** and again at **17:48 CEST**, each carrying the app name, the Hungarian sentence and the dashboard link. The CUSTOMER channel is **SKIPPED**, which is correct. **I asserted an absence without looking at the instrument** — the hub's own Events and Notifications tabs show all of it — and I did it by reasoning from a memory note (`lxc-docker-oom-signals-unreliable`, R-528) instead of reading the hub. **That is R-628's shape exactly, four days on, and from the same hand.** **WHAT IS ACTUALLY WRONG, and it is narrower and real:** `notifier.go:715-726` keys the alarm on `container|startedAt` in an `oomSeen` map and emits **once per container lifetime**. So **six hours of continuous thrashing — 4,530 worker kills — produced exactly ONE mail**, severity `warning`, never escalating, while the app went on reading `running`. **A six-hour storm is indistinguishable from a single transient kill.** The hub's App Telemetry panel did carry the magnitude (RomM: **5,023 errors, 632 warnings**, peak 855 MB) but nothing turns that into a second, louder signal. So the fix worth having is not a detector — there is one — but an ESCALATION: a warning that repeats for hours should stop looking like a warning that happened once. **AND THE LESSON FOR R-462's METHOD, which is the real cost:** every `proven` verdict in the update night and in the twenty-eight measures the app for the **minutes of the walk**, not for a day of running. `romm` passed its walk, was seeded, read back and restored — and broke two hours later. **`proven` currently means "the update applied and the data survived", NOT "the new version runs".** Needs: decide between rolling the catalog back to 5.0.0 and raising romm's `mem_limit` (measured, not guessed); and a soak longer than a walk before any future promotion. **FIXED AND MEASURED 2026-09-22, in two steps, and the FIRST step was still a guess.** *Step 1 (operator's choice):* the limit was raised 512M → **768M** (`app-catalog-felhom.eu@886956d`). It slowed the kills from ~12/min to ~7/min and **stopped nothing** — 37 SIGKILLs in five minutes, `OOMKilled` still true, and the cgroup's own `memory.peak` read **exactly 768 MiB**: it hit the new ceiling and died there. *Step 2, from a MEASUREMENT instead:* the per-process RSS inside the container reads **~216 MiB per warm uvicorn worker**, so the image's default of four workers plus the master needs **~882 MiB** before nginx and the job runner — more than any sensible limit for this box. **The lever was in the image all along:** `/init:143` runs `--workers "${WEB_SERVER_CONCURRENCY:-4}"`. **Four workers is a SERVER default on an appliance serving one household.** Setting **`WEB_SERVER_CONCURRENCY=2`** (`app-catalog-felhom.eu@f4eb94f`, lLine truncated
|
||||
| **R-723** | Monitoring & notifications | P3 | **[P3-LOW] A fresh box sends the operator two mails on day one that describe nothing wrong.** MEASURED 2026-09-29: `Operator email sent for tester-1/node_recovered` 2 s after the new box's first controller report (the customer's previous box had been silent 12 days — the new box is not a recovery); and `backup_tier_skipped (warning) — Whole-guest backup tier felhom-pbs skipped: its storage does not exist on the host (never provisioned or removed)` → operator mail at 19:27 UTC, 7 min after enrolment, because the first whole-guest run fired before the off-site tier's descriptor arrived (applied ~19:35 UTC, backed up fine at 19:37). Operator-only, so no household is alarmed, but an operator learns to ignore both. **Fix direction:** `node_recovered` not for a host enrolled < N min ago; the skipped tier inside the first hour after enrolment is `info`, not a mailed warning. **FIXED 2026-09-30 (hub v0.126.0):** no `node_recovered` when the customer's host was enrolled after the outage began; `backup_tier_skipped` in a box's first hour recorded, not mailed. Real recoveries and old boxes' skips still mail (controls). Unit + red-proof (RP40, RP41); the live proof is Tester-2's first hour. | **WATCHING** (2026-10-03 triage: the row's verdict was finished, but it names open work no other row carries — FIXED in hub v0.126.0; the live proof is Tester-2's first hour) — **FIXED — hub v0.126.0; live proof at the first real install; owner: CC** | — | — | CC |
|
||||
| **R-724** | Monitoring & notifications | P3 | **[P3-LOW] The status pages disagree with each other in small ways a household notices.** MEASURED 2026-09-29 on a fresh box: Beállítások reads „Mentés ütemezés 02:30 / 03:00" while Biztonsági mentés says 02:30 / 03:30 / 04:15 / 04:30–08:30; Beállítások shows a raw `2026-09-29T19:22:30Z`; its „Helyi cím (LAN)" and „Átjáró" read „nem állapítható meg" on the page the household is asked to read out for remote help; the backups overview still says „Következő mentés — 0 órája" (the age of the LAST run, under the word "next" — noted 2026-09-14, never filed); the dashboard shows the backup at 19:33 where the backup pages say 21:33 (R-500, still reproducing). **FIXED 2026-09-30 (controller v0.283.0):** Beállítások shows the real schedule and a local time; the dashboard's last backup is local time (R-500); „az előző N órája készült" replaces „0 órája". **NARROWED — remaining:** „Helyi cím (LAN)" / „Átjáró" read through the file-sharing container, so a box without it says „nem állapítható meg" — not a text fix (the read needs another path). | **NARROWED — the LAN/gateway read only; owner: CC** | — | — | CC |
|
||||
| **R-177** | Monitoring & notifications | P4 | **There is no operator-triggerable "run the fill check now" path.** `fill-watch` is reachable only on its daily 03:30 schedule plus the once-at-startup run added in controller v0.191.1 — so the only way to exercise it on demand is to restart the controller | **READY (S) — NEW 2026-08-02** | — | **Noticed while live-validating R-167 on 9201, not by a failure.** It cost a controller restart per observation during validation, and it costs the same on a support call: after a customer frees space, nobody can confirm the warning has cleared without restarting their controller or waiting until 03:30. **Partially mitigated already** — v0.191.2 makes every run log a positive observable (`checked N filesystem(s), M unreadable/skipped, K notification(s)`), so at least a run that DID happen is visible; the gap is triggering one. The scheduler has `GetJobs` but no run-now, so this is a general affordance, not a fill-watch one — **scope it as "run a named scheduler job now", operator-gated.** **ID established free:** `grep -ro "R-177\b" documentation/ *.md` → 0 hits | CC |
|
||||
@@ -411,7 +395,7 @@ stopping line that lies.
|
||||
| **R-793** | Business & legal | P4 | **[P3-LOW] Enterprise / BUSL code ships inside four open images — Cal.com and Docmost (EE folders, off without a key), Outline (BUSL-1.1: no commercial "Document Service"), meilisearch v1.36 in Wanderer (EE modules).** READ 2026-10-02 (`audits/licences-2026-10-02/TABLE.md`). Each is fine as the catalog runs them: no EE key, the household's own Outline is not a Document Service, Wanderer uses plain search. **Watch:** never turn on an EE feature, never switch Karakeep's/Wanderer's meilisearch to the `-enterprise` image, and re-read on each major. | **WATCHING — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: a watch item; nothing is wrong as the catalog runs them.** | — | — | CC |
|
||||
| **R-794** | Business & legal | P4 | **[P3-LOW] redis 7.4 (RSALv2 / SSPL, not OSI) runs as a private cache in seven apps: dawarich, docmost, immich, nextcloud, outline, paperless-ngx, romm.** READ 2026-10-02 (`audits/licences-2026-10-02/TABLE.md`). Read as permitted (a private cache only its app uses is not Redis offered as a service — inferred). Valkey (BSD-3) or redis 8 (AGPL option) removes the question. **Needs:** a ladder step per app to valkey or redis 8, through the harness — no hurry. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: the row itself says no hurry; usage read as permitted.** | — | — | CC |
|
||||
|
||||
## Process & tooling — 88 rows (P3 4, P4 84)
|
||||
## Process & tooling — 75 rows (P3 4, P4 71)
|
||||
|
||||
| ID | Category | Sev | What | State | Blocked on | Next action | Owner |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
@@ -425,16 +409,13 @@ stopping line that lies.
|
||||
| **R-162** | Process & tooling | P4 | **`docker diff` is the gate's only witness, and its failure mode is quiet.** The gate's power comes from `docker diff` excluding mounted paths, which makes "in the writable layer" mechanically decidable — an implementation detail of the overlay driver. On a driver where `docker diff` is unsupported or lies, the gate degrades to the mount-occupancy and writability legs **and would not say so**. | **WATCHING** — a limitation, not a defect | — | It **fails closed**: the canary self-test would stop reporting BROKEN and the gate would then refuse to report at all. What is wrong is the message — it would blame the prober rather than the driver. Revisit only if a non-overlay storage driver ever ships | CC |
|
||||
| **R-169** | Process & tooling | P4 | **CI can only report, because there is no gate in the road.** Every felhom repo pushes straight to `main` with no pull request, so there is no merge for a status check to stand at. R-168's runner therefore notices a broken push *after* it has landed | **WAITING-ON-OPERATOR** (a working-style decision, not a defect) | an operator ruling | Making CI *blocking* requires two things this task deliberately did NOT do, because both change how the operator works and that is not a task's call: **(a)** branch protection on `main`, and **(b)** a pull-request workflow instead of direct-to-`main` pushes. The cost is real — every change would need a PR, which for a single-operator project may be worse than the disease. **The current arrangement is two nets, and it is not nothing**: `.githooks/pre-push` REFUSES locally, and R-168's runner NOTICES when that hook was skipped or was never armed in a clone, and emails. The honest gap is the window between a `--no-verify` push landing and the operator reading the alarm. Decide only if that window ever actually costs something | operator |
|
||||
| **R-206** | Process & tooling | P4 | **The build-cache cap and the weekly prune exist only as a hand-edited `/etc/docker/daemon.json` on DooPlex — not in Ansible, so a rebuild loses them.** The `node_housekeeping` role must also carry the prune, which today it is forbidden to run | **READY (M) — NEW 2026-08-05** | — | **The spike validated the recipe; this row builds it.** Three parts. **(a) Template `/etc/docker/daemon.json`** with the **`policy` array** form — **the flat form (`{"gc":{"reservedSpace":…}}`) is SILENTLY IGNORED**, measured: the daemon starts, logs nothing, and `docker buildx inspect` still reports the built-in defaults. **The oracle is `docker buildx inspect`, never `dockerd --validate`** — the validator returned `configuration OK` for a bogus key AND for the config that then **crashed the daemon** (`filter` takes one value per policy entry, not an array; `error initializing buildkit: filters expect only one value`). **(b) Narrow the role's Docker ban** (`node-housekeeping.sh.j2:10-14`) to permit exactly `docker builder prune -af` and nothing else — the ban's stated premise ("Docker here runs only unrelated jarr-* dev containers") is obsolete: the growth is Felhom Go build cache. **The measured prune is SYNCHRONOUS** (150.35 GB back at t+0, two consecutive polls <1 MB apart within 60 s) — **unlike containerd's image GC, so it needs no `settle_imagefs` equivalent**, but it MUST measure the filesystem rather than trust the command: `prune` claimed **156.9 GB** and the filesystem returned **150.35 GB**, the 6.5 GB gap being layers still shared with images. **(c) A restart-safety note in the role:** a bad `daemon.json` takes the daemon down AND leaves the `unless-stopped` dev containers stopped — they needed a manual `docker start` — so the role must restart-and-verify, not validate-and-assume. Recipe + every measurement: `audits/SPIKE-dooplex-buildcache-2026-08-05.md` | CC |
|
||||
| **R-207** | Process & tooling | P4 | **`DRY_RUN=1` on `node-housekeeping.sh` is NOT non-mutating — it destroys the metric history it is supposed to let you inspect** | **READY (S) — NEW 2026-08-05** | — | `write_metrics()` (`node-housekeeping.sh.j2:119-150`) has **no `DRY_RUN` guard at all** — `DRY_RUN` appears in it only inside a log line — and it is called from an **unconditional `EXIT` trap** (`:150`). A dry run therefore atomically renames over the live node_exporter textfile, overwriting `node_housekeeping_last_success_timestamp_seconds` with *now* and `reclaimed_bytes` with ~0 — **resetting the staleness clock `HousekeepingStale` watches and erasing the 8-week reclaim history**. Confirmed by reading the source in both the 2026-08-05 audit and this spike; **neither run executed it**, so the history survives. Fix: guard `write_metrics` on `DRY_RUN`, or have the trap skip it. Pairs naturally with R-206 (same file, same role) | CC |
|
||||
| **R-208** | Process & tooling | P4 | **Every Felhom Go build re-downloads its modules because `ARG VERSION` sits ABOVE the module-download layer — ~440 MB of dead cache per build, 90.5 GB of the 157 GB** | **READY (S) — NEW 2026-08-05, ROOT CAUSE PROVEN** | — | **Measured, not inferred.** All **208** retained `go mod download` records carried **`Usage count: 1`** — not one was ever reused in ~a month of builds. The mechanism was isolated by four controlled builds: an unchanged tree rebuilt with the **same** `--build-arg VERSION` → `RUN go mod download` **CACHED**; the **same** tree with a **new** `VERSION` → **executed**. `COPY go.mod ./` stays CACHED either way, which is the tell: a `COPY`'s key is content-based, while a `RUN`'s key includes the stage **environment**, and `ARG VERSION`/`ARG GIT_COMMIT` are declared *before* the download in `felhom-controller/controller/Dockerfile`. Since every real build passes a fresh version, the layer is invalidated **every single time**. **`felhom.eu/hub/Dockerfile` has the identical defect** (`ARG VERSION`/`ARG BUILD_TIME` above `COPY go.mod go.sum*` → `RUN go mod download`) — and because both Dockerfiles produce byte-identical `buildx du` description strings, the 208 records are a COMBINED count and must not be attributed to one project. **Fix shape (one line each, not applied here):** move the `ARG VERSION`/`ARG GIT_COMMIT`/`ARG BUILD_TIME` declarations down to just above the final `go build`. **Worth more than the cap and the move combined** — the cap bounds the symptom, this removes the source. `build.sh`'s `rm -rf` + `cp -a` and its host-side `go mod tidy` were **ruled out by fingerprinting**: the tree is byte-identical across runs and `tidy` is a no-op | CC |
|
||||
| **R-209a** | Process & tooling | P4 | **The SSD2 move has NOT survived a reboot, so by this project's own standard it is not fully validated** | **WATCHING — NEW 2026-08-05** | the next DooPlex reboot | **Operator ruled explicitly: do NOT reboot DooPlex.** Uptime verified unbroken (7 weeks 6 days, since 2026-06-10). **The distinction is stated rather than glossed: the MECHANISM is proven** — the guard is wired into both units and containerd refuses to start when a required mount's device is absent — **but the CONSEQUENCE is not**: that a real boot mounts `/mnt/ssd_2` before containerd starts, in this host's actual ordering. Mount-ordering reasoning is precisely the class this project has been burned by (`RequiresMountsFor` RE-MOUNTS rather than refusing — the ep0 lesson), and `CLAUDE.md` prefers a consequence assertion over a mechanism one. **Two deliberate consequences: (1)** the rollback copy `/var/lib/containerd.pre-move-2026-08-05` (**34.3 GB on `/`**) **STAYS** until a reboot validates — which is why `/` sits at 54% and not lower; deleting it now would trade a cheap 34 GB for the only cheap way back. **(2)** validation is **automatic and needs no one to remember it**: `felhom-store-postboot-check.service` (oneshot, enabled, dry-run PASS at install) runs at **every** boot and writes `RESULT: PASS`/`FAIL` to `/var/log/felhom-store-postboot-check.log`, asserting positively that `/mnt/ssd_2` is mounted, that containerd's root is on it, that **`/var/lib/containerd` does NOT exist** (the empty-store trap), that ≥100 images are visible and that both dev containers run. **Next action: after the next reboot — planned or not — read that file; on PASS, `rm -rf /var/lib/containerd.pre-move-2026-08-05` returns ~34 GB to `/`** **P3's prune already removed the urgency: `/` went 86% → 53% used and SSD1's Longhorn disk went `Schedulable=False (DiskPressure)` → `Schedulable=True` (18.85% → 50.32% available).** The move was ruled "cap then move"; the cap is in and the pressure is gone, so this is now a deliberate choice rather than a rescue. **The numbers, measured (`Crucial-SSD-240G`, `/mnt/ssd_2/data/longhorn`, `storageMaximum` 235,148,750,848):** available today **214,958,080,000 (91.41%)**; 25% floor **58,787,187,712**. Moving the whole containerd tree at steady state (~31.5 GB images + ≤30 GB cache ≈ 65 GB) leaves **63.77%, i.e. +38.8 pp above the floor — comfortably safe as measured.** **But `storageScheduled` on SSD2 is 139,586,437,120 while `df` says only 20,094,939,136 is actually used** — Longhorn has overcommitted 6.9× — and if those volumes ever inflate to their scheduled size, the same disk lands at **13.00%, i.e. 12 pp BELOW the floor → `Schedulable=False`**, which is exactly the failure that just took SSD1 out. **Recommendation: do the move only together with setting `storageReserved` on SSD2 to cover the containerd tree (~80 GB); SSD2 reserving zero while HDD2 and HDD4 each reserve 500 GB is an anomaly in its own right.** Mechanism, if it goes ahead: **containerd's `root` in `/etc/containerd/config.toml`** (the key is present but commented out) — **not** Docker's `data-root`, which would move only 0.62 GB. Guard: `RequiresMountsFor=/mnt/ssd_2` on `containerd.service` **and** `docker.service`, remembering that **`RequiresMountsFor` RE-MOUNTS rather than refusing** ([[ep0-datastore-volume-move-2026-07-27]]) — so it must be tested with a genuinely absent device, and the move is not validated until it has survived a **reboot**. Full pre-analysis: `audits/SPIKE-dooplex-buildcache-2026-08-05.md` §P6 | operator + CC |
|
||||
| **R-210** | Process & tooling | P4 | **Which of the 345 local images may be deleted — 131 controller tags and 62 hub tags exist ONLY on this box and are not recoverable by `docker pull`** | **WAITING-ON-OPERATOR** — NEW 2026-08-05 | an operator ruling | **Nothing was deleted; this is a list, not an action.** The registry was queried directly: `felhom-controller` has **76** tags in Gitea vs **207** locally, `felhom-hub` **45** vs **107**. The **131 + 62 local-only tags are all OLD** — controller `0.39.0`–`0.135.0` plus `v0.35.0`–`v0.39.0`, hub `0.9.0`–`0.57.0` plus `v0.7.2`–`v0.13.0` — while everything from controller `0.136.0` and hub `0.58.0` upward IS in the registry and therefore re-pullable. **Size the prize honestly before spending a decision on it:** per-tag sizes sum to 139.29 GB, but that double-counts shared layers — `docker system df` puts the **real** dedup'd image footprint at **31.02 GB with 27.02 GB reclaimable**, i.e. an order of magnitude less than the build cache P3 already returned. `docker image prune -a` would remove 343 of 345 (only `redis:7-alpine` and `postgres:16-alpine` are held by running containers). **CC's view: not worth doing for the space** — it buys ~27 GB against 199 GB now free, and its only real benefit is dropping unrecoverable clutter | operator |
|
||||
| **R-230** | Process & tooling | P4 | **Three instruction/memory follow-ups deliberately left by the part-2 session (2026-08-06), each needing a decision rather than an implementation.** (a) **A ruling is owed on auto-written staleness.** The hand-written `CLAUDE.md` files are now clean of version literals and expired blocks — the gate enforces it — but `MEMORY.md`, which Claude writes and which is the LARGER half of what loads (8.4k tokens vs the root file's 6.6k), carries **21 lines with component version literals**, **5 with bare host addresses**, and an entry still reading *"demo boxes REMOTE till ~08-02"* — the same expired-TEMPORARY class the gate was built to kill, now surviving in the one file the gate's content rules do not cover. **Partly actioned 2026-08-06 (close-out), and the ruling is STILL OWED:** the **three statements that were actively false** were corrected — `R-193 decision open` (closed 2026-08-05), `demo boxes REMOTE till ~08-02` (the box answers on the home LAN), `OPEN R-25b` (shipped 2026-07-21) — and gate check 6 now **WARNs** on version literals, host addresses, expired statements and stale-open citations in the index. WARN, never FAIL: Claude writes that file between sessions, so a hard failure would refuse a human's push over a line no human typed, and the warning is read by the model that will next edit it. **The remaining 32 version literals and 4 host addresses were deliberately left** for that loop. What is still owed is the bulk-correction ruling. **Correcting the premise:** the earlier report's "three expired statements" were all FALSE POSITIVES — each matched an ISO date inside a markdown link target, i.e. a filename — while the one real expired claim carried no ISO date at all. (b) **CLOSED 2026-08-06 (close-out)** — the workspace-root `CLAUDE.md` **is now a relative symlink** to the versioned copy, so the divergence class is gone rather than policed. Check 5 learned two shapes: for a link it asserts the target resolves to a real file (**a dangling link is worse than a diverged copy — the instructions load NOTHING and there is no content left to notice is wrong**), for two files byte-identity as before, so a clone elsewhere is unaffected. **Proven, not assumed:** three fresh sessions logged `session_start` for the link path, and a fourth **with no tools at all** quoted standing rule 1 verbatim — the content reaches the model, not just the path. (c) **The spec-as-failing-test pilot**, approved in principle and not started (was R-229(d)). | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-263** | Process & tooling | P4 | **C7 — „This is the ONLY writer of `StoragePath.BackupTarget`" is false, and nothing pins it.** `settings/settings.go:1317-1319`, on `SetBackupTarget`. **`ClearBackupTarget` (`:1358-1363`) also writes the field**, 17 lines below, in the same file. **The GUARANTEE the comment protects is intact and that is why this is filed small:** the sentence continues *"registration must never set it (E-2 §3: a drive never acquires a role by appearing)"*, and `ClearBackupTarget` only ever writes `false`, so no path other than `SetBackupTarget` **grants** the role. **What is wrong is the claim as written, and the absence of anything holding it:** `backup_target_role_test.go` exercises the behaviour and asserts nothing about writer uniqueness, so if a third writer appeared tomorrow — one that granted — the comment would still read as settled and the suite would still be green. **This is the class's own definition:** an invariant asserted in prose with no test pinning it. **Fix shape:** one word (*"the only writer that GRANTS the role"*) plus a test that fails when a second granting writer appears. **Method honesty:** found in a sample of **60 of 2652** production invariant comments — the "is the only" form only, chosen because a uniqueness claim is the one form a grep can falsify. **~2592 production and all 1440 test invariant comments were NOT examined**, so C7 has the weakest coverage of the seven classes and the task's instruction to check the tests' own claims is **owed, not discharged** | **READY** — owner Viktor | — | — | operator |
|
||||
| **R-286** | Process & tooling | P4 | **A control drawn from the same channel as the measurement cannot detect a defect in that channel — and this one passed while the measurement was wrong.** The P7 check asked *"did the hub record anything?"* against a stale snapshot, got "no", and then validated itself with *"is the hub recording ANY events today, for anyone?"* — **against the same stale snapshot**. It answered "2 events all day", which was internally consistent and entirely false. The standing rule (*an absent log line is not evidence*) was followed in form: a positive control WAS run. **It was the wrong kind of control**, and nothing in the rule as written says so. **The independent channel existed and was available the whole time: the operator's mailbox.** One glance at it would have shown eight alarms in the window. **The durable lesson, to be added where the standing rules live:** a control must come from a DIFFERENT channel than the measurement — same query, same snapshot, same API, same clock all fail this. **Concrete follow-through owed:** (a) add this to the standing rules in `runbooks/workspace-CLAUDE.md`; (b) any hub-state check in a runbook must copy `-wal` or query the pod directly, never `cat hub.db` alone — the trap `operations/nodes.md` already documents | **READY (S) — NEW 2026-08-09** | — | Parent: R-281 (withdrawn) | CC |
|
||||
| **R-287** | Process & tooling | P4 | **`felhom-agent` CI is red for a TRUE reason, and the diagnosis it was filed under is wrong in every particular.** The task premise was *"the gate is sensitive to being run against a tag ref rather than a branch"*. **It is not.** `check-published-versions.py` enumerates releases from the **Gitea tags API** (`/api/v1/repos/admin/felhom-agent/tags?limit=200`, `main()`), so the checked-out ref is irrelevant; and the two previous tag pushes **passed** (run 190 `v0.126.0`, run 216 `v0.127.0`). **What is actually true:** run 267 (main, `28ba8593b8`, 2026-08-08 14:29 UTC) printed `ok v0.120.0: binary downloadable`; run 284 (tag, **the same commit**, 2026-08-09 09:30 UTC) printed `FAIL v0.120.0 — binary NOT downloadable (HTTP 404)`. **A published release became uninstallable between those two runs.** The registry now holds exactly the ten newest versions (0.121.0…0.128.0); `0.128.0` was published **2026-08-08 16:47 CEST = 14:47 UTC, eighteen minutes after run 267**, and `0.120.0` is gone. **WHO REMOVED IT IS NOT ESTABLISHED, and that is stated rather than guessed:** `package_cleanup_rule` is **empty** (queried in Postgres), `app.ini` sets no package limit, `publish-agent.sh` only pre-deletes the version it is publishing (`:77`), the Gitea pod has **53 days uptime and 0 restarts** so `RUN_AT_START` did not fire, and I wrote that *"no DELETE on the packages API appears in 48 h of Gitea router logs"*. **THAT SENTENCE WAS AN OVER-CLAIM AND IS WITHDRAWN 2026-08-09 (evening).** Re-checked: `kubectl logs --since=72h` on the Gitea pod returns nothing older than **2026-08-09 16:35** — the container log has rotated, and it contains **zero** `api/packages` lines even for requests I made myself. **The log does not cover the window, so its silence was never evidence** — the exact rule this project keeps re-learning. **A SECOND ATTEMPT TO ATTRIBUTE THE DELETION ALSO FAILED, and the deleter remains NOT ESTABLISHED.** Sources exhausted: (1) no register row records a package prune — **R-210** is the only prune-adjacent row, it is `WAITING-ON-OPERATOR`, it says in terms *"Nothing was deleted; this is a list, not an action"*, and it concerns **local Docker images on DooPlex, not the Gitea registry**; (2) `package_version` has **no soft-delete column** (`created_unix, creator_id, download_count, id, is_internal, lower_version, metadata_json, package_id, version`) so a deletion leaves no row; (3) Gitea's `action` feed carries **no package operation at all** in 2026-08-08 → 2026-08-09, and **nothing whatever on the evening of 2026-08-08**; (4) a uniform *"newest ten per package"* cap is **not visible in the current state** — `felhom-agent` generic holds 10, but `felhom-controller` and `felhom-hub` container packages hold **19 each**. ~~**It may be unestablishable from this side: Gitea keeps no package-deletion trail.**~~ **ESTABLISHED 2026-08-10, AND IT WAS WRITTEN DOWN ALL ALONG — INSIDE R-267.** The execution record is in the R-267 row (the Configuration-page performance row), because pruning artifacts is what made that page fast: *"Pruned to the newest 10 per package on the operator's rule, with the live-vouched golden/agent/floor asserted into the KEEP set before a single DELETE was issued; 33 deletions, all HTTP 204, and golden 0.210.0 / agent 0.128.0 / agent 0.127.0 verified still fetchable afterwards."* **Every corroboration checked and every one holds:** the arithmetic (23 agent + 7 golden = 30, plus three older agent versions 0.81.0/0.80.0/0.79.0 that *"only became visible after the first 30 deletions moved them onto page one"* = **33**); and the live PAGINATED listing — 14 pages, 653 package-versions — showing `felhom-agent` generic at **exactly 10** and `felhom-golden` generic at **exactly 10**, which is what a newest-10 prune leaves. **The midnight-cleanup candidate is RETIRED.** **AND MY OWN COUNTER-ARGUMENT WAS WRONG, IN THE WAY R-267 WARNS ABOUT.** On 2026-08-09 I argued against the prune because *"container packages hold 19 each"*, so no uniform cap was visible. That number came from an **unpaginated** query, which the API caps at 50 per page. Paginated, the containers hold **270** and **169** — they were never in the prune at all, and the generics are at exactly 10. R-267 records the identical trap one paragraph above the sentence I could not find: *"An unpaginated listing is not evidence of a total — this repo's own rule, walked into while measuring."* I walked into it a day later and used the bad number to argue against the record that documents it. What remains established is unchanged — v0.120.0 downloadable 2026-08-08 14:29 UTC, 404 by 2026-08-09 09:30 UTC, and the generic package now holding exactly the newest ten. **THEREFORE NO GATE WAS SILENCED AND NO WORKFLOW WAS CHANGED.** Silencing it would hide a released-but-uninstallable version, which is the exact R-115 defect the gate exists to catch. **It will recur:** if the ten-version window is real, Line truncated
|
||||
| **R-288** | Process & tooling | P4 | **The capability map is too long to be read, and that is why it stops being true.** `architecture/00-capability-map.md` is **134 642 bytes / 19 456 words across 99 table rows in only 159 lines** — because the rows ARE the length. Measured, longest first: the unaided-recovery-journey row is **3 024 words**, the offsite-password-recovery row **1 087**, the unattended-restore-proof row **971**, the app/guest-network-failure row **904**. That single longest row is a novella of nested corrections, each appended rather than resolved. Its own verification stamp reads **2026-07-16 against evidence corpus @ felhom.eu tip `4b18cc5`** (line 23) — three weeks stale, which is the measurable consequence: nobody re-reads a row they cannot finish. **This is the project's memory, so restructuring it is surgery and wants daylight** — filed, deliberately not attempted in the 2026-08-09 session. **What the shape should probably be:** one line of status per capability plus a dated evidence pointer, with the argument moved to the audit it came from | **READY (M) — NEW 2026-08-09** | — | Do not fold this into another session; it needs its own. **SECOND CONCRETE COST, 2026-08-10 — and it is a different failure mode from the first.** An *execution record* — 33 package deletions on the operator's rule — was undiscoverable for two days because it lives **inside the row about the Configuration page being slow**. Two sessions searched for it: one reported "no register row records a package prune", the other exhausted the Gitea logs, the activity feed and the schema before concluding it might be unestablishable. It was in `OPEN-ITEMS.md` the whole time. **The first cost (2026-08-09) was two records that looked contradictory and were not; this one is a record that could not be found at all.** Illegibility now has two measured costs and they are different in kind: prose rows make claims ambiguous, and rows-about-other-things make facts unfindable. **The rule this earns is in CONTEXT.md: a record that lives inside a row about something else has not been recorded** | Viktor |
|
||||
| **R-289** | Process & tooling | P4 | **R-182's register row describes a defect the code no longer has — an OPEN row that is a false alarm.** The row reads *"A full disk tells the operator about ONE app and silently swallows every other app's refusal for an hour"*, cited at `hub/internal/notify/dispatcher.go:268`. **Read against live source 2026-08-09, that is fixed:** the per-run digest `backup_run_failures` is allowlisted (`hub/internal/api/handler.go:1837`), operator-only (`dispatcher.go:423`) and templated (`notify/templates.go:48`); `recovery_unit_capture_failed` is now a **record-only** event (`dispatcher.go:376`) whose notification IS the digest, listing every failed app in one mail; and a cooldown drop now writes a `suppressed` row instead of vanishing (`dispatcher.go:314-330`). The capability map already records the fixed shape (*"EVERY failing app, in ONE mail per run"*). **So the register is behind the code, which is the mirror of the decay this session was looking for** — the session expected stale PROOFS and found a stale DEFECT. **Not closed here, deliberately:** the digest's *delivery* has never been observed end to end (the page's own "an app crashes — the email leg has never been confirmed" card), so the honest move is to re-scope R-182 to that residue rather than tick it | **READY (XS) — NEW 2026-08-09** | — | Re-scope R-182 to "the digest has never been seen delivering", or close it and open that | CC |
|
||||
| **R-290** | Process & tooling | P4 | **Most capability-map rows that back a green dot cite no evidence document at all — measured, 20 of 28 probed.** The page's *Walked* means *"done end to end on real hardware, evidence on file"*. Extracting the evidence column for the 28 rows behind the page's claims found a `tests/` or `audits/` path in **8**; the other 20 carry prose only. **Consequence, applied this session:** of 32 claims the page drew as Walked, **12 were downgraded to Built** because no walk document exists for them — `install.installer-by-tag`, `use.lifecycle`, `drives.enrol`, `drives.migrate`, `backup.tier1`, `backup.whole-machine`, `backup.restore-proof`, `fault.selfheal`, `fault.operator-email`, `fail.drive-filling`, `fail.lost-recovery-code`, `fail.hub-down`. **This is not a claim that those twelve are false** — several are near-certainly fine — it is a claim that nothing on file distinguishes them from an opinion, which is exactly what the status word promises. **The gate now enforces it going forward:** `scripts/check_stands.py` fails on `status: walked` with no `evidence:` source. **What is owed:** either a walk document per row, or an honest demotion in the map itself (the map is the source; the dataset only follows it) | **READY (M) — NEW 2026-08-09** | R-288 | The dataset was corrected; **the capability map itself still says PROVEN-LIVE for these rows** and is the thing to fix | Viktor |
|
||||
| **R-291** | Process & tooling | P4 | **CI's installability assertion is now BOUNDED by a retention number, and the narrowing is recorded here so it can be widened deliberately rather than discovered.** `check-published-versions.py` demanded that **every** `v<semver>` tag still be downloadable while the registry demonstrably does not retain every version — two sensible rules that cannot both hold, which is why CI went red at a commit whose own run had been green the day before, and would have gone red again at the next publish. **The fix couples them:** `felhom-agent/scripts/retention-policy.json` is THE number (`generic_versions_kept: 10`) and the check reads it. **WHAT CI NO LONGER COVERS, stated plainly: a released version older than the retention window is no longer asserted downloadable.** Its git TAG and its config tree are still asserted — only the binary's presence is dropped — and the check **prints the dropped versions on every run**, so the narrowing cannot go quiet. Controls run: widened to 11 the evicted version re-enters and convicts (exit 1); the policy file removed gives INCONCLUSIVE (exit 2), never silently unbounded. **The number is an OBSERVED state, not a located ruling** (R-287) and the file says so. **The better bound, recorded rather than built:** the hub's vouched `min_agent` floor — nothing can install an agent below it, so a sub-floor version being un-downloadable costs nothing real; it needs the gate to read the hub, which is network it does not have today | **READY (S) — NEW 2026-08-09** | R-287 | ~~Widen or replace the number when the deleter is established~~ — **CONDITION RELEASED 2026-08-10: the deleter IS established (R-287), so the operator is no longer blocked on establishing what was already written down.** The number can now be confirmed or replaced on its merits. The better bound remains the vouched `min_agent` floor | CC |
|
||||
| **R-315** | Process & tooling | P4 | **The wire-contract gate's positive control FAILS on the new wire: it checks name-presence, not decodability.** R-311 declared `hub -> agent (GET /escrow/retained)` as a fourth ROOT, and the gate's tag count rose 174 → 182, so the fields ARE inspected. But renaming the agent-side `superseded_at` json tag to `superseded_at_RENAMED` **still passed** — because the string `superseded_at` also occurs as a map key in the agent's local-API response, and the check is a repo-wide name search. The gate documents this ("name-reachability is not use"), so it is a known limit rather than a regression — but it means **declaring this wire bought documentation, not enforcement**, and a report that claimed coverage would have been wrong. The mutation was asserted to have applied before the run | **READY (M) — NEW 2026-08-12, RANK 3** | R-311 | Make the check resolve the RECEIVER'S mirror type and compare field-by-field, or state per-root which kind of check it got. **A gate whose positive control fails is an instrument nobody has calibrated** | CC |
|
||||
@@ -444,9 +425,7 @@ stopping line that lies.
|
||||
| **R-346** | Process & tooling | P4 | **`ActiveEnterTimestamp` answers a different question than the one a slope measurement asks, and on ep0 right now it is wrong by 5 h 56 m.** Found while taking R-341's first dated check. ep0's `proxmox-backup-proxy` has `MainPID=551655` started **2026-08-18 09:51:04Z** (`ps -o lstart=`), but `systemctl show -p ActiveEnterTimestamp` reads **03:54:54Z** and `NRestarts` reads **0** — because the 4.2.5-1 upgrade **re-exec'd** the daemon rather than restarting the unit, so systemd never observed a stop. Anyone anchoring "when did this proxy generation start" on `ActiveEnterTimestamp` would divide 388 descriptors by 52.1 h instead of 46.2 h and report **178/day instead of 201.6/day — ~12% low** — while every field consulted looks healthy and consistent. **This is the workspace rule's own case, in a new place:** ask of a timestamp *what exactly must have happened for this to be set?* Here the answer is "the unit entered active", which is not "this process started". `NRestarts=0` is the tell, and it reads like reassurance. | **READY (XS) — NEW 2026-08-20** | — | R-341's command already uses `ps -o lstart= -p $MainPID` and is correct; the risk is a future reader "improving" it to a systemd property. Add the reason as a comment beside that command in the R-341 row (done), and check whether any other slope or uptime check in the repo or in `scripts/felhom-tenantsync.sh` anchors on a systemd timestamp where it means a process start. | CC |
|
||||
| **R-364** | Process & tooling | P4 | **Accented-text search is an instrument that silently transforms its input, and discipline alone has failed at least three times.** (1) 2026-07-20, `ssh → pct exec → bash -c`, nearly a wrong "banner cleared" claim (`felhom-controller/.claude/rules/ui-hungarian.md:19-22`). (2) 2026-08-13, `kubectl exec … sh -c grep` returned **0 for three strings that were present**, one step from a wrongly-reported failed hub deploy. (3) 2026-08-21, `tar -tf` rendered `őszibarack.md` as `\305\221szibarack.md`; recording the fixture's name bytes from that listing would have been wrong. **NOTE: that is two inside two weeks plus the founding case a month earlier — a third inside the two-week window is not on record.** | **OPEN — LOW** | — | **PROPOSED, NOT BUILT:** a helper that refuses to report a zero for any pattern containing a byte ≥ 0x80 unless a negative control also returns zero AND an ASCII anchor known to be present returns non-zero. Three probes, one helper, no judgement at the call site — because judgement is what failed. | CC |
|
||||
| **R-374** | Process & tooling | P4 | **Three C1 refusal cases were judged borderline, left unfiled, and never named — so nobody can re-open the judgement.** `audits/CAMPAIGN-12-class-sweep-2026-08-08.md:123`: *"'Names a route' is a judgement, not a predicate — two readers could disagree on the borderline cases, and three of the 19 were called borderline and left unfiled."* The disclosure is honest and is exactly the right thing to write; **what is missing is WHICH three.** An unnamed borderline case cannot be re-judged by a second reader, which is the only remedy a judgement call has. **Age when filed: 14 days.** | **OPEN — LOW** | — | Name the three in that document, or file them as one row listing them. No code. | CC |
|
||||
| **R-376** | Process & tooling | P4 | **The placement decision that cost four mis-filed defect reports was never recorded as a decision anywhere, and the architecture folder's own marker convention lives in one document of eight.** Established 2026-08-22 by reading, not by citation: the hot/bulk split exists as a single unmarked bullet at `architecture/01-topology-and-trust.md:150-152` — **no dated entry in the decision log, no `R-` row, no recorded date for when it was taken.** Meanwhile its CONSEQUENCE (40 of 53 templates declare no path) is stated as **[FACT]** at `07-backup-architecture.md:296-299`. A reader met a marked observation beside an unmarked choice and reasonably asked whether it should be so — four times (R-370). **Measured marker usage 2026-08-22:** `[DESIGN]`/`[FACT]` appear in **1 of 8** architecture documents (`07`: 10 and 35 uses); the other seven had **zero**. **Done this session:** the legend is carried into all seven, each stating explicitly that an unmarked statement means *not yet classified* and never *observed*; the hot/bulk bullet is marked `[DESIGN]` with an honest pointer saying the decision has no original date on record; and a decision-log entry was written in `CONTEXT.md` **to give it a home, not to claim it was decided then**. **Deliberately NOT done:** the existing statements in those seven documents were not swept into one marker or the other — that is a large judgement exercise and a wrong mark is worse than none. | **OPEN — MEDIUM** **Re-ranked 2026-10-03: MEDIUM→P4: documentation hygiene; no household meets it.** | R-370 | Mark statements as sessions touch them, per the template's §4 map. Do not bulk-classify. If the original date of the hot/bulk decision is ever recovered, put it in the log entry. | CC |
|
||||
| **R-377** | Process & tooling | P4 | **`CONTEXT.md`'s standing rulings are 188 KB in one section with no sub-headings, and that is why nobody reads them.** Measured 2026-08-22: `CONTEXT.md` is 217 KB, of which **187,913 bytes — 86% — is a single `## Standing rulings` section** carrying 39 `S-` ids and 153 bullets under **one** heading. **This session deliberately did NOT compress or split it**, and the reason is the ruling itself: `PROMPT-TEMPLATE.md` §3.4 and this file's own contract say the decision log is *dated, never edited afterwards*, and it is the only place that answers *"has this been proposed before, and why did we say no?"* — compressing it destroys exactly that. With no per-ruling delimiter, any mechanical split risks cutting a live ruling from its reason, which is the failure this whole arc is correcting. **So the problem is navigational, not volumetric, and the fix is structural: give each ruling a sub-heading with its `S-` id and date.** Then it can be linked, cited and found without a single word being edited. **13 mentions of SUPERSEDED already sit inside that blob** and cannot be separated from live text safely today. | **OPEN — LOW** | R-369 | Add per-ruling sub-headings only. Do not compress, do not reorder, do not edit any ruling's text. | CC |
|
||||
| **R-390** | Process & tooling | P4 | **The golden-bake runbook omits `pveam update`, and the failure it produces names the wrong cause.** `documentation/runbooks/RUNBOOK-manual-build.md` §4.1 step 2 says to list the current Debian template because "the exact point release rots" — but on the drill VM's `virgin` snapshot **the `pveam` INDEX is stale too**, so `pveam available` offers an old point release and `pveam download local <that>` fails with **`400 Parameter verification failed. template: no such template`**. That reads as a typo or a bad argument, not as an old index, and it costs a diagnosis every time. **Hit on two consecutive bakes** (golden 0.222.0 and 0.223.0, both 2026-08-23). The runbook is otherwise correct verbatim — the qemu launch line, the token-read-inside-the-VM pattern and the acceptance markers all worked unchanged. | **OPEN — LOW** | — | Add `pveam update` as its own numbered step before the listing, and say WHY: a snapshot that never changes carries an index that never updates, so the rot warning already in the step applies to the index as well as to the release. Recorded meanwhile in the workspace memory `golden-bake-needs-pveam-update` and in `documentation/tests/golden-0.223.0-2026-08-23/README.md`. | CC |
|
||||
| **R-391** | Process & tooling | P4 | **Gate 11 (observations) is registered in three of the four runners; `app-catalog-felhom.eu` is the exception.** The controller and agent runners already carried a shared-gate mechanism (`SHARED_REUSE`, `SHARED_INSTRUCTIONS` pointing into `felhom.eu/scripts/`), so registering there was one constant and one `GATES` line each. **`catalog_gates.py` has no such mechanism:** its `run_gate` joins every entry against its OWN `scripts/` directory, so it cannot invoke a sibling repo's script at all; and its loop appends `--all` to every gate unconditionally, which the observations gate would read as a path. Registering there therefore needs `run_gate`'s contract widened AND the argument handling changed — a refactor of a runner whose shape is deliberately different (per-app scoping, network/runtime gates excluded from `--fast`), in a repo this task marked out of scope. **The exposure today is nil** — `app-catalog-felhom.eu/REPORT.md` has no observations section, and the gate passes quietly on that — but a future catalog session could write one and nothing would read it. **Filed rather than left as a sentence in a report, which is the exact failure R-389 records.** | **OPEN — LOW** | — | Either give `catalog_gates.py` the `SHARED_*` absolute-path mechanism the other two runners already have and stop appending `--all` to gates that do not take it, or state in that repo's CLAUDE.md that its REPORT.md carries no observations section by convention. **Do not copy the gate script** — the shared checker lives in ONE place (`felhom.eu/scripts/`) and copying it is the drift the shared pattern exists to prevent. | CC |
|
||||
| **R-392** | Process & tooling | P4 | **No architecture document covers the two-AI workflow.** `documentation/architecture/` holds eight documents and **all eight cover the product** — topology, host agent, control-plane authorization, hub, off-site connectivity, backup, controller modules, capability map. Nothing records how the Claude.ai / Claude Code split works, what each side owns, how skills and `.claude/rules/` are scoped, or why. **The absence was found by trying to fill the template field, not by a survey:** the task that added the five process skills (2026-08-25) had to name an owning architecture document and could not, and the template requires that be recorded rather than passed over. The exposure today is low — the split is stable and both sides work — but it lives entirely in the operator's head and in chat, which is precisely the shape of a commitment nothing enforces. | **OPEN — LOW** | — | Write one architecture document for the agent-tooling layer: which AI owns which artifact class (`TASK-*.md`, `RUNBOOK-*.md`, validation), how skills are scoped and installed, what belongs in a `CLAUDE.md` versus a skill versus a rules file, and the reasoning for each boundary. **The rules themselves already exist** in `skills/felhom-doc-authoring/SKILL.md`; what is missing is the map of who owns what. Do not restate the doc-authoring rules there — point at that skill. | CC |
|
||||
| **R-393** | Process & tooling | P4 | **A decision-log skill for unattended runs was considered and deliberately deferred.** Filed 2026-08-25 by the session that added the five process skills, so the deferral is a decision on the record rather than a thing that was dropped. **The gap it would close:** an overnight or unattended run makes dozens of decisions and the operator can only reconstruct them by reading the whole transcript, which is exactly what nobody does. The proposal is an appended row per decision — what was chosen, why, the evidence pointer, and the result — so a long run is reconstructable in a page. **Why it was NOT built with the other five:** the other five are text files that need nothing but the existing installer glob. This one needs a helper script to append rows and a storage convention for where the log lives and when it is rotated, which makes it an implementation task with its own acceptance criteria, not a skill file. | **OPEN — LOW** | — | Decide the storage convention FIRST — most likely a per-session file beside the session's evidence directory, never `REPORT.md`, which is overwritten every session (the R-341 shape). Then the skill, then the helper. **Check it does not duplicate `felhom-handoff`**, which already owns the end-of-session note; a decision log is the during-the-run half and the two must point at each other rather than overlap. | CC |
|
||||
@@ -460,13 +439,10 @@ stopping line that lies.
|
||||
| **R-424** | Process & tooling | P4 | **`one_register_gate.py`: a real defect parked under the roadmap state `idea` is invisible to it.** MEASURED 2026-09-01 with a correctly-shaped 5-column row. This is **declared** in the gate's own docstring as residual hole 1 — *"the state column is a human judgement, and a defect written under `idea` looks exactly like a proposal to this gate"* — so it is honest, not hidden. Recorded here because a hole declared only in a docstring is not in the register, which is this project's own standing rule. No cheap fix: distinguishing a defect from a proposal mechanically is the thing the gate cannot do. | **OPEN — declared, not hidden; recorded so it is not re-derived** | — | — | CC |
|
||||
| **R-425** | Process & tooling | P4 | **`offbox_rename_gate.py` scans a fixed three-entry `FILES` list.** MEASURED 2026-09-01: `NAS-mentés` in a new `backups_offbox_extra.html` passed. The scope was correct when written and silently narrows every time the feature grows a file. Fix: scan the offbox feature's files by pattern, or assert the FILES list against a discovered set so a new file fails until it is classified. | **OPEN** | — | — | CC |
|
||||
| **R-426** | Process & tooling | P4 | **The decoy-coverage exemption list — 20 registered gates that ship WITHOUT a decoy test, each named.** `scripts/decoy_coverage_gate.py`'s `EXEMPT` map is debt, and this row owns it so it lives in the register and not only in a Python literal. **Four kinds:** (a) genuinely covered in the 2026-09-01 sweep but not yet moved into a suite — `hub-copy`, `instructions`, `docker-v`, `image-pins`; (b) blocked by an open hole and therefore un-assertable as rejecting — `site` (R-423), `one-register` (R-424), `offbox-rename` (R-425); (c) shared scripts whose decoy lives in `felhom.eu` and is counted there — `reuse-refs`, `instructions`, `observations` in the controller and agent runners; (d) **no plausible decoy constructed yet** — `hostinstall`, `wire-contract`, `due-checks`, `published`, `image-resolvable`, `volume-persistence`. Group (d) is the honest unknown: six gates whose soundness is UNTESTED, not established. **The list is green today and shrinks; a NEW gate with no decoy fails immediately.** | **OPEN — 20 names; group (d) is six untested gates** | — | — | CC |
|
||||
| **R-427** | Process & tooling | P4 | **`closed_register_gate.py` checks ONE direction only: an open word in a CLOSED row. The mirror — a CLOSED verdict on a row still sitting in `OPEN-ITEMS.md` — is unchecked, and there are TWELVE.** MEASURED 2026-09-01 during the decoy sweep, by reading the leading verdict of every open row with the gate's own predicate: **R-385, R-387, R-341, R-378, R-405, R-88a, R-88b** read unambiguously closed; **R-123, R-190, R-352** read `PARTLY CLOSED` / `MITIGATION SHIPPED` and almost certainly belong where they are. **The rows were NOT moved by this session** — telling a finished row from a partly-finished one is a judgement, and R-378 is itself the record of what happens when a machine makes that judgement on a substring (six still-open rows moved out of the register). **This is R-405's finding mirrored:** that row exists because R-87 sat in the CLOSED file while its state read READY, and the gate written for it looks only the way it was bitten. Fix: the same leading-verdict predicate applied to `OPEN-ITEMS.md`, reporting rather than convicting until the twelve are adjudicated by a person — a gate registered while twelve rows fail it would refuse every push. | **OPEN — 12 rows named; the adjudication is Viktor's, the gate is mine** | — | — | CC + operator |
|
||||
| **R-437** | Process & tooling | P4 | **The register compression sweep is OWED, and it was deliberately NOT run inside the 2026-09-01 beta-line session — this row is the record of that choice, not a note.** **The ask:** compress what has closed in `OPEN-ITEMS.md`. **The measurement, taken before deciding:** 181 rows, 316 KB of row text, of which **12 rows / ~25 KB (about 7 %) carry a CLOSED/DECIDED/ANSWERED leading verdict.** So the sweep buys little and touches everything. **Why it was refused as a side-task, and the citation matters:** a compression sweep is the exact operation that has already gone wrong here. The 2026-08-22 sweep (`ef6ac6f`, R-376..R-378) matched a status word ANYWHERE in the row, moved rows that were not closed, and **R-378 caught six in the same session and missed a seventh** — R-87 sat in the wrong register for nine days while the ranking paragraph pointed at nothing (R-405). **That is a session-scale hazard, and running it as the tail end of a session about something else is how it happened the first time.** **WHAT IS OWED, scoped so it can be picked up cold:** (1) classify by the **LEADING VERDICT** of the state cell only — the rule `closed_register_gate.py` already implements and red-proofs, never a whole-row match; (2) move, never rewrite — a compressed row that loses its evidence is worse than a long one; (3) run `closed_register_gate.py` before and after and quote both; (4) re-read the ranking paragraph afterwards, because that is the surface that silently went stale last time. **Not urgent:** the file is 688 lines and every gate reads it in well under a second. | **OPEN — owed; needs its own session, not a tail end** | — | — | CC |
|
||||
| **R-454** | Process & tooling | P4 | **[P3-LOW] Five `internal/web` test files have been `gofmt`-unclean for an unknown length of time, and nothing notices.** MEASURED 2026-09-02: `gofmt -l controller/internal/web/` reports `backups_split_test.go`, `claim_code_naming_test.go`, `disk_health_test.go`, `r400_debug_routes_test.go`, `recovery_test.go` — at the **baseline** commit `960d29b0612c`, i.e. not introduced by v0.233.0 (both files added that day are clean). **`go vet` does not check formatting and `controller_gates.py` has no formatting gate**, so the only thing that would ever surface this is someone running `gofmt -l` by hand, which is how it was found. **Not reformatted in the same session, deliberately** — the minimal-changes rule, and a five-file whitespace commit inside a feature release makes that release's diff unreadable. **Small, and the cost of NOT having the instrument is the row:** the count can only grow, and every future `gofmt -l` run produces noise that hides a real one. Fix is two lines: a `gofmt -l` gate in `controller_gates.py` plus one formatting commit, in that order (the gate first, so the commit is provably complete). Owner: **CC.** | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: code-formatting tooling only; no household meets it.** | — | — | CC |
|
||||
| **R-457** | Process & tooling | P4 | **[P3-LOW] A test that hardcodes a date AND asserts an age derived from it is green on the day it is written and red the next morning — one instance PROVEN, six candidate files named.** MEASURED 2026-09-03: `TestGroupD_BadgeRendersOnBothSurfaces` (shipped the previous day in v0.233.0) pinned a fixture `catalog_since: "2026-07-18"` and asserted the rendered string `"Frissítés elérhető — 46 napja"`. **The pure badge tests inject a clock; the RENDER test does not and cannot** — it goes through the production templates, which call the funcmap entry `updateBadge`, which reads `time.Now()`. The suite was green on 2026-09-02 and **FAILED on 2026-09-03** with *"the behind badge is missing"* on both surfaces, because the true answer had become 47. **Fixed by DERIVING the fixture** — `catalog_since` is computed as *today minus 46 days*, so the test asserts the real number through the real clock and cannot rot. **THE CLASS, which is why this is a row and not just a fix:** a clock-reading test that also carries a date LITERAL is a bomb with a fuse of unknown length, and the suite being green is not evidence it is defused — it is evidence the fuse has not burned down yet. **NAMED AS UNCHECKED CANDIDATES, NOT ACCUSED** — six other test files contain both a `20xx-xx-xx` literal and `time.Now()`: `internal/backup/offbox_test.go`, `internal/web/handler_export_upload_test.go`, `internal/web/r103_tier2_action_test.go`, `internal/web/dashboard_backup_card_test.go`, `internal/web/async_restore_test.go`, `internal/stacks/installed_test.go`. Mixing the two is not itself a defect — it is one only where a literal feeds an assertion evaluated against the real clock — so each needs reading, which is a sweep and not this session. **The instrument that would end the class:** run the suite once under a faked future date in CI and see what turns red. Owner: **CC.** `felhom-controller` v0.234.0 CHANGELOG | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: test-suite hygiene; no household meets it.** | — | — | CC |
|
||||
| **R-488** | Process & tooling | P4 | **[P3-LOW] `go test ./internal/backup` takes 5½ minutes: 89 off-site tests wait on real clocks.** MEASURED 2026-09-13 (`-v` timings, run alone: 581 tests, 333 s in total, 89 of them ≥ 1 s — `TestOffbox*`, `TestOffbox3a*`, `TestOffboxRun*`, `TestR4xx*` reconstitute fixtures at 3–8 s each). The controller's per-commit gate is therefore ~6 minutes, most of it sleeping, and two concurrent runs of the package looked like a hang. **Fix shape:** the waits are `waitForHealthy`-style polls and retry back-offs with fixed durations; make them seams the fixtures shorten (the R-457 rule: one clock). Not a correctness defect. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: developer speed only.** | — | — | CC |
|
||||
| **R-492** | Process & tooling | P4 | **[P3-LOW] `cfg.Paths.HDDPath` is empty on every box and still has readers; delete it.** R-465 audited its six readers and found every one falling back; R-490 (v0.242.0) gave the last one, `systemInfo`, the same fallback. The global now carries no information on any box and its deletion was deferred twice. **Fix shape:** remove the field, its env binding and the readers' fallback branches; a build proves nothing reads it. Next controller release. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: dead-code cleanup; every reader already falls back correctly.** | — | — | CC |
|
||||
| **R-501** | Process & tooling | P4 | **[P3-LOW] The documented "confirm your CI run" recipe reads only the LAST page of the jobs list, and that list is not in id order — so it can report a run as missing that exists and passed.** MEASURED 2026-09-14 for felhom.eu commit `38848ff`: `actions/jobs?limit=1` → `total_count 335`; the recipe's page `T/50+1` = page 7 held ids 473…570 and **no match for 8 minutes**; a scan of all seven pages found the job on **page 6** (ids 380…581): `job id=581 name=gates status=completed conclusion=success completed_at 2026-09-14T14:18:23Z`. Pages are not sorted (page 1 ids 1…273, page 2 84…208). `actions/tasks` listed the same run first (`id 582, run_number 335, success`). **Fix shape:** the recipe in `felhom.eu/CLAUDE.md` (end-of-session checklist) scans every page and matches `head_sha`; say so in the same sentence that warns about the id offset (R-417). | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: operator/CC tooling recipe only.** | — | — | CC |
|
||||
| **R-502** | Process & tooling | P4 | **[P3-LOW] The bootstrap regression harness is run by NO gate and NO CI — and it had never exercised the pairing banner.** MEASURED 2026-09-14: `grep -rn bootstrap-modes scripts/*.py .gitea/workflows` → nothing; the harness is run by hand in `felhom-iso-assistant:trixie`. While adding the R-496 checks, its fake hub's `register` reply turned out to carry no `pairing_code`, so `print_pairing_banner` returned early in every run: the banner that a household reads first had no test at all until ISO v1.27.0. **Fix shape:** register the harness in `repo_gates.py` behind a container-available check that reports INCONCLUSIVE (never skip-as-pass) when docker is absent, with a decoy. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: test harness coverage; no household meets it directly.** | — | — | CC |
|
||||
| **R-507** | Process & tooling | P4 | **[P3-LOW] The proof-install harness cannot drive the graphical installer, so a release's graphical entry is proven only up to its password screen.** MEASURED 2026-09-14 on VM 332 (ISO 1.27.0): `qm sendkey 332 tab` did not move focus (both password copies landed in one field), `mouse_move 1237 772` + `mouse_button 1` did not move the cursor or press Next, while `alt-n` did advance a page. The TUI entry is fully drivable. The gate's "proof install on BOTH menu entries" was met for 1.26.1 (by a person) and not for 1.27.x. **Fix shape:** measure QEMU `input-send-event` with absolute coordinates, or a VNC client on DooPlex; until then a release's graphical proof is an operator click-through. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: release-test tooling; an operator click-through covers it.** | — | — | CC |
|
||||
| **R-551** | Process & tooling | P4 | **[P3-LOW] No Tier-0 box can put the escrow ceremony in the state R-546 fixes — paused AND connected to its agent — so the readiness branches are proven only by tests.** FOUND 2026-09-17 while live-validating controller v0.246.0 (R-546). The branches (reminder bar held back while the agent's preflight is not `ok`; the waiting card on `/backup/escrow`; `POST /api/escrow/start` refused 409 before staging) need a box whose off-site tier is configured, whose escrow is NOT done, and whose controller reaches the agent. Measured on demo-hp: **9201** reaches the agent but is escrowed (the bar is off by design there; making it paused would be hand-set state on the standing demo box, and a real ceremony would supersede its live escrow — the hub keeps ONE `host_escrow` row per host, `host_id PRIMARY KEY`); **9202** is paused-capable but has **no local-API token at all** (its `bootstrap.json` holds only `schema`, `customer.id`, `disposition`), so its readiness is always UNKNOWN and the bar always shows. The fresh-bind window where this state occurs naturally (~17 min, chaos night Phase 0) needs a fresh install. **What IS proven:** five tests driving the real pages and handler through `ServeHTTP` with a fake agent, each red-proofed; and chaos night measured live that the agent's preflight is red for ~17 minutes after a bind and turns green by itself (`evidence-chaos-night-2026-09-17/phase0-escrow-*`). **Fix shape:** give the scratch guest a local-API token through the agent's own provisioning path, or walk R-546 on the next fresh install. | **READY - rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3->P4: test-coverage comfort; behaviour is unit-tested and no household meets it.** | — | — | CC |
|
||||
@@ -480,14 +456,11 @@ stopping line that lies.
|
||||
| **R-588** | Process & tooling | P4 | **[P3-LOW] ISO release records live in two different places, so "was the gate run for this image?" cannot be answered by looking.** FOUND 2026-09-18: the release records for 1.27.0 and 1.27.1 are directories under `documentation/tests/iso-release-<ver>-<date>/`, and there is none for **1.28.0** — which led me to conclude its gate had not been run. It had: the record is `documentation/audits/evidence-backup-promise-2026-09-16/phaseD-iso-gate.txt`, inside an audit about something else. **The gate's own rule is "a criterion with no recorded observation is a criterion that was not run"**, and that rule is unenforceable while the records have no single home — the question it answers has to be settled by a full-text search for a checksum, which is what it took here. **Fix shape:** one home, `documentation/tests/iso-release-<ver>-<date>/`, and a line in `iso-release-gate.md`'s Result-recording section naming it; optionally a check that every published ISO version has a record directory. | **READY - rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3->P4: documentation hygiene.** | — | — | CC |
|
||||
| **R-591** | Process & tooling | P4 | **[P3-LOW] `Stack.Copy()` is a deep copy with one shallow field, and the field is new.** FOUND 2026-09-20 while adding the catalog's English overlay: `controller/internal/stacks/manager.go` `Copy()` deep-copies `Meta.DeployFields` (with nested `Options`), `Meta.OptionalConfig` (with nested `Fields`), `Meta.Integrations`, `Meta.HealthCheck` and `Meta.InitialCreds` — and does NOT copy the new `Meta.I18n` map, which the struct assignment leaves shared between the original and the "copy". **It is safe TODAY and that is exactly the shape worth filing:** `Metadata.For` reads the overlay and never writes to it (pinned by `TestForDoesNotMutateTheReceiver`), so nothing can observe the sharing yet. The hole is in the CONTRACT — a function whose whole purpose is "a snapshot the caller may mutate" now has a field that is not one, and the next person to write through an overlay will find a bug with no failing test in front of it. **Fix shape:** deep-copy `I18n` in `Copy()` and pin it with a test that mutates the copy's overlay and asserts the original is unchanged. Alternatively state in `Copy()`'s comment that `I18n` is deliberately shared and immutable, and pin THAT with a test. Either is fine; silence is not. | **READY - rank P3-LOW; owner: CC (controller)** **Re-ranked 2026-10-03: P3->P4: latent contract gap, safe today.** | — | — | CC |
|
||||
| **R-594** | Process & tooling | P4 | **[P3-LOW] The catalog copy gate can CONVICT a retrieval promise but has no way to REGISTER a true one.** FOUND 2026-09-20 translating batch 3 (R-560 slice 5). Vaultwarden's invite step and its sign-up setting both ended „…can open an account", and the English retrieval-promise pattern reads `can … open` as the claim that sealed backups can be opened. **The conviction was a FALSE POSITIVE** — opening an account is not opening a backup — and the two sentences were reworded to „can sign up", which is also the better copy, so nothing is blocked today. **The gap is structural.** The shared vocabulary this gate copies (`scripts/customer_copy_vocab.py`) states the design explicitly: these stems are NOT banned, because each carries a claim that is sometimes TRUE, and *"an occurrence must be REGISTERED with a reason in the consuming gate's allowlist"*. The hub gate has `ALLOWLIST_EN`; `app-catalog-felhom.eu/scripts/check-copy-i18n.py` has none, so the only ways past it are to reword or to bypass the gate — and a catalog app whose English genuinely says a file can be restored (a backup app, a versioned document store) has no honest third option. **Fix shape:** an `ALLOWLIST_EN` of `(app, path, reason)` in the gate, a decoy proving a REGISTERED occurrence passes and an unregistered one still convicts, and a check that every entry still matches something (a stale allowlist entry was R-299's shape, and the retrieval gate has gone red on stale entries before). | **READY - rank P3-LOW; owner: CC (catalog)** **Re-ranked 2026-10-03: P3->P4: gate tooling; nothing blocked today.** | — | — | CC |
|
||||
| **R-602** | Process & tooling | P4 | **[P3-LOW] The language a signed-in page uses is NOT the language a cookie asks for, and a live probe that forgets this reports a fixed defect as unfixed.** FOUND 2026-09-21 verifying R-598 on guest 9201. `GET /backups` with `felhom_lang=en` returned the **Hungarian** page. That is correct — `langFor` step 2 says a request carrying a session reads the household's saved setting and deliberately ignores the visitor cookie, so a signed-in family never sees a language a previous visitor picked on the sign-in page of the same browser — but it means **the cookie is the right instrument for the anonymous claim/login/bind pages and the wrong one for every page behind auth**, where `?lang=` is. A session that had run only the cookie probe would have concluded R-598 was still open and fixed it a second time. **This is a documentation gap, not a code defect**, and it is the kind that costs a whole session: nothing in `10-localisation.md` §2.2 or in any runbook tells a prober which instrument to use where. **Fix shape:** four lines in `10-localisation.md` §2.2 — a table of surface → language instrument — and a pointer from the live-validation section of the workspace rules. Recorded meanwhile in `audits/i18n-closing-2026-09-21/live/backups-page.md`. | **VERIFY** (2026-10-03 triage: 10-localisation.md now has the surface-to-language table and records the cookie-vs-?lang= finding — felhom.eu/documentation/architecture/10-localisation.md:500-512 and 809-815) — **READY — rank P3-LOW; owner: CC (docs)** **Re-ranked 2026-10-03: P3->P4: documentation for probers only.** | — | — | CC |
|
||||
| **R-603** | Process & tooling | P4 | **[P3-LOW] An English string containing an apostrophe silently never matches on a rendered page, and a `strings.Contains` assertion reads exactly like a missing sentence.** FOUND 2026-09-21 while writing the R-598 render tests. `backup.target.absent` was first written as *"The system backup's drive cannot be reached…"*; `html/template` escapes `'` to `'`, so the page carried the sentence and every assertion for it failed. **The failure mode is the expensive part:** the test said *"the English absent-drive copy never reached the page"*, which is indistinguishable from the handler not being wired — and the obvious next move is to go and re-fix the handler. Reworded to avoid the possessive, and all 23 new English values were then swept for `' " < > &` (zero). **The Hungarian bundle has never hit this** because Hungarian copy uses „quotes" and few apostrophes; **English copy will hit it again.** **Fix shape:** either a bundle gate that refuses an HTML-escapable character in a value destined for a page (and an allow-list for the ones that legitimately need one), or a test helper that compares against `html.EscapeString(want)` so the assertion cannot be fooled. The gate is the better shape — the helper only protects tests that remember to use it. | **READY — rank P3-LOW; owner: CC (controller)** **Re-ranked 2026-10-03: P3->P4: test tooling.** | — | — | CC |
|
||||
| **R-605** | Process & tooling | P4 | **[P3-LOW] A catalog gate that REFUSED TO RUN and a gate that ran and could not decide print the same word, so a reader cannot tell which happened.** FOUND 2026-09-21 while answering why the chaos night's update round could not run. On 2026-09-17 `check-image-resolvable` and `check-volume-persistence` both returned INCONCLUSIVE and the drawn `update` action was replaced with `use` (`audits/DRILL-chaos-night-2026-09-17.md:692-695`). **Neither script is defective — they behaved exactly as designed**, and both headers say why: a detector that cannot prove itself must refuse to report rather than guess (`check-image-resolvable.py` cites the 2026-07-21 incident where a Docker Hub throttle read as 24 of 65 pins falsely dead). **What is missing is the DISTINCTION.** `check-volume-persistence.py`'s `self_test` refuses to evaluate ANY app when it cannot build its canary image — a harness-level refusal — while `classify()` returns a per-app UNDETERMINED for an app that wrote nothing; `check-image-resolvable.py` likewise separates a harness-level canary failure (exit 2 at `check()` L180-183) from a per-pin throttle (L121-128). **`catalog_gates.py`'s VERDICT map collapses all of them into one `INCONCLUSIVE` label**, so the operator-facing summary cannot say whether the gate ran at all. **The cost is real and already paid:** no raw stdout of the 2026-09-17 run survives in either evidence directory, so the exact triggering path is INFERRED from the code plus the documented throttle precedent, not observed — a distinct summary line would have recorded it for free. **Fix shape:** have each gate's exit distinguish "the harness refused" from "the result is undetermined" (a third exit code, or a marker line the runner matches), and have `catalog_gates.py` print the two differently. **Ships with a decoy each way (R-421): a run whose canary fails must NOT read as a per-app undetermined, and vice versa.** Small. | **READY — rank P3-LOW; owner: CC (catalog)** **Re-ranked 2026-10-03: P3->P4: tooling report clarity.** | — | — | CC |
|
||||
| **R-617** | Process & tooling | P4 | **[P3-LOW] The Gitea API token this project uses for pushes cannot create a repository through the documented endpoint, but CAN through `repos/migrate` — so "the token cannot do it" was nearly recorded as a fact when the truth was "one endpoint refuses it".** FOUND 2026-09-21 creating the drill catalog. Both `~/.git-credentials` tokens carry `write:misc,write:notification,write:package,write:issue,write:repository`; `POST /api/v1/user/repos` requires `write:user` and answers **403**, and `POST /api/v1/admin/users/<u>/repos` requires `write:admin` and answers 403 too. `POST /api/v1/repos/migrate` with the same token answered **201** and created the private repository. **Why this is a row and not a note:** a session that stopped at the first 403 would have recorded "CC cannot create a Gitea repository" — an unfalsifiable capability claim of exactly the shape the workspace's standing rule 2 forbids — and every later drill would have been designed around a limit that does not exist. **What it needs:** one line in the operations notes saying which endpoint to use, and (optional, operator) a token scoped for the job so the migrate route is not load-bearing. Evidence: `audits/update-night-2026-09-21/03-drill-repo.txt`. | **READY — rank P3-LOW; owner: CC (docs)** **Re-ranked 2026-10-03: P3→P4: a note for CC's own tooling; no household meets it.** | — | — | CC |
|
||||
| **R-624** | Process & tooling | P4 | **[P3-LOW] Three of the catalog's apps cannot be seeded by ANY headless route, and for two of them that is a deliberate security decision — so the upgrade harness has a permanent ceiling nobody has written down.** FOUND 2026-09-21 while widening R-462 from 3 apps to <N>. **`vaultwarden` and `zipline` close self-registration ON PURPOSE** — vaultwarden by `SIGNUPS_ALLOWED=false` (R-512, *„a stranger who guesses vault.<domain> must not be able to register"*), zipline by answering `E1037: User registration is disabled` — and neither ships a CLI that could make an account instead. So there is no route to a first account without the admin secret, and **that is correct**: the harness must not be the reason a customer-facing app accepts strangers. **`gitea` is a different and fixable case:** the template sets no `INSTALL_LOCK`, so a fresh instance sits in its web-installer state and `gitea admin user create` refuses (`MustInstalled() [F] Unable to load config file for a installed Gitea instance`); POSTing the installer form first would work and was simply not written tonight. **Why this is a row rather than three notes:** `09` §3 decision 6 says the upgrade test goes to **all** apps, and R-462 is costed as if every app is reachable given enough fixture work. It is not. There is a class — *apps whose only account-creating route the catalog deliberately closes* — for which the honest maximum is `inconclusive` unless the harness is given the app's admin secret at deploy time, which is a decision nobody has taken. **Needs:** the class named in R-462's scope so the remaining count is honest; a decision on whether the harness may hold an app's admin secret (it already holds the ones IT generates — see R-619); and, separately and cheaply, a gitea installer-form fixture. Evidence: `audits/update-night-2026-09-21/apps/{vaultwarden,zipline,gitea}/verdict.json`. **— NIGHT 2026-09-23:** `gitea` stayed inconclusive on both venues (its installer); `vaultwarden`/`zipline` not attempted (closed sign-up, by design); `code-server`, `outline`, `rallly` have no front-door seed route (listed, not moved); `bentopdf`, `glance`, `crafty-controller`, `wger`, `wanderer` (meilisearch) and `uptime-kuma` have no fixture tonight (listed, not moved). **-- 2026-09-30: the ceiling was WRONG for three of the six named apps.** outline has a front-door first-run route (`POST /api/installation.create` — its self-hosted setup, refused once a team exists), rallly signs up through its own better-auth API with the e-mail code read from its own table in place of a mailbox, and zipline 4's first-run route is `POST /api/setup` (the fixture had tried `/api/auth/register` and `/api/auth/setup`, which are not it). Fixtures for outline and rallly are in `upgrade_fixtures_box.py` and both apps moved on both venues; zipline's fixture now tries `/api/setup` first (measured on the bench: a SUPERADMIN made, the login works). **What remains in the class:** vaultwarden (closed sign-up by design, no CLI), code-server (a browser IDE). `audits/pg-last-six-2026-09-30/`. **-- 2026-09-30 (evening): gitea's fixable case is DONE** — the fixture posts gitea's own first-run installer form (the form's own defaults + an admin; no CSRF on that form), measured on the bench and through the setup gate on 9202; gitea 1.27.0 → 1.27.3 published `d7ba60c`. What remains in the class: vaultwarden, code-server. `audits/more-night-apps-2026-09-30/` | **READY — rank P3-LOW; owner: CC (catalog harness)** **Re-ranked 2026-10-03: P3→P4: an upgrade-test coverage limit; no household meets it.** | — | — | CC |
|
||||
| **R-652** | Process & tooling | P4 | **[P3-LOW] The memory watch counted the kernel's file cache as the app's memory.** MEASURED 2026-09-23 night on the bench: nextcloud 34.0.4 read **100 %** of its 1 GiB and immich's PostgreSQL **100 %**, each with **0** kernel `oom_kill`s — `memory.peak` includes page cache, which the kernel drops before it kills anything. Under the watch as built (R-635 follow-up) both would be marked `memory_tight`, and the gate would demand a raised `mem_limit` — a customer-box capacity figure — for cache. **Done the same night (09 §3 decision 22, CC — operator may reverse):** the watch samples the app's own memory (`anon` of `memory.stat`) every 15 s; the mark and the ladder's `memory_peak_pct` read it where measured; the cgroup peak stays beside it (`memory_cgroup_peak_pct`). **Open:** romm's backfilled entry carries M1's 80.9 % cgroup peak (measured before the anon sample existed) — re-measure it on its next step; and decide whether an app whose anon is low but whose cgroup stays pinned at its limit (cache thrash) should be marked at all. Evidence: `audits/night-2026-09-23/apps/nextcloud/bench-1024M/`, `apps/immich/bench-noanon/`. | **READY — P3; owner: CC (catalog harness)** **Re-ranked 2026-10-03: P3→P4: the main fix is done; what is left is harness tuning, no household meets it.** | — | — | CC |
|
||||
| **R-693** | Process & tooling | P4 | **[P3-LOW] The memory watch marks a Node app `memory_tight` at any limit — its heap sizes itself from the limit.** Measured 2026-09-25 on the bench (docmost 0.96.0, harness v4): the app's own memory (`anon`) peaked at **349 MB of 384 MB (90.9 %)**, then, with the limit raised to 512 MB, at **431 MB of 512 MB (80.4 %)** — 0 OOM kills and 0 restarts in both 10-minute watches (~12 000 requests each). So the mark (decision 22's "does not fit the memory") fires for an app that fits, and the gate's remedy (raise the limit) cannot clear it. docmost moved with the limit raised to 512 MB (decision 39). **Needs:** a basis that tells growth-to-fill from pressure (e.g. kills/restarts plus a GC-pressure signal, or a second watch at a higher limit showing the peak scales), or a per-app `memory_scales_with_limit` fact. `audits/night-2026-09-26/C/bench-run1/`, `…/bench-run2/` | **OPEN — P3; owner: CC** **Re-ranked 2026-10-03: P3→P4: a harness judgement problem; no household meets it.** | — | — | CC |
|
||||
| **R-705** | Process & tooling | P4 | **[P3-LOW] There is no way to run the night's chain now — only its pieces.** Asked by the operator 2026-09-28 (to finish a proof in the day). What exists (read from source, controller v0.278.0): the backup page's off-site run-now (`POST /backup/offbox/run`) runs the dump leg first (the R-44 pre-phase: DB dumps, volume dumps with brief app stops, unit capture) and then the push — used live on demo-hp 2026-09-28 15:17, 3m57s; the debug API has `backup/dbdump`, `backup/crossdrive` (Tier 2), `backup/integrity`, `backup/offsite-proof`. **Missing:** the automatic update leg (`RunUpdateLeg`, chained only to the scheduled off-site job) and the whole-guest backup (the agent's, on its own 24 h / 7 d cadence) have no manual trigger; the only way to run the chain in order is to move the backup window (`POST /backups/window`), which takes W..W+2h at least. **Needs:** a debug action "run tonight's chain now" (dump → Tier 2 → off-site → update leg, in order, one at a time), and an agent-side "whole-guest backup now" for demo boxes. Not built. **-- 2026-09-28 evening: the controller half BUILT (v0.279.0):** debug `POST /api/debug/backup/night-chain` runs dump → Tier 2 → off-site → update leg in order, one at a time, refused while anything else runs; the leg gets its normal length from its own start. Proven on 9202: 44 s, a second press 409, the leg deadline 22:00. **Still open: the whole-guest backup (agent side) has no manual trigger.** | **OPEN — P3, agent half only; owner: CC** **Re-ranked 2026-10-03: P3→P4: operator/test comfort; the controller half is built.** | — | — | CC |
|
||||
| **R-731** | Process & tooling | P4 | **[P3-LOW] The catalog-currency comparison cannot see an upstream that changes its TAG SHAPE, and two newest tags need a release check.** MEASURED 2026-09-30 (`audits/catalog-currency-2026-09-30.md` §2 item 5): the same-shape rule read three apps as up to date that were not — gramps-web (`v25.6.0` → upstream dropped the `v`, at `26.9.1`), jellyfin (`10.11.11` → two-part `12.1`), kimai (`apache-2.57.0` → plain `2.67.0`; the plain tag's digest equals `apache`'s, so kimai moved on 2026-09-30). A control pass over every shape caught them. Not checked: whether `mariadb:13.0` and `gitea/gitea:28.0.0` (pushed 2026-09-30 00:15 UTC) are general releases. **Needs:** the currency script's shape-switch control made standing (it is in the audit's tools today), and the two release checks before either is walked. **-- 2026-09-30: the two release checks answered** (`audits/immich-first-start-2026-09-30/D/D2-release-checks.txt`): **gitea v28.0.0** is a general release (GitHub: not prerelease, not draft, 2026-09-29) — upstream renumbered 1.27.x → 28; **mariadb 13.0** is `Stable` but a short-term `Rolling` line (no EOL date), while 12.3 and 11.8 are the LTS lines — so a move of any of the four MariaDB apps to 13.0 would leave LTS. Not done: the shape-switch control made standing. | **NARROWED 2026-09-30 — only the standing shape-switch control is left; owner: CC** **Re-ranked 2026-10-03: P3→P4: only a standing check in a tooling script is left.** | — | — | CC |
|
||||
| **R-739** | Process & tooling | P4 | **[P3-LOW] The test bench cannot run wanderer at all: its web server calls the database at the PUBLIC name `https://<SUBDOMAIN_DB>.<domain>`, which the bench has no name or TLS for.** MEASURED 2026-09-30 (more-night-apps brief, Part B): on bench 9401 the catalog template (v0.20.0) came up `wanderer-db` healthy, `wanderer-search` healthy, `wanderer` **unhealthy** for 12 min, every page 500 („Error 0: Something went wrong"); `PUBLIC_POCKETBASE_URL=https://hike-db.gate.invalid`. So the harness can only ever answer `inconclusive` at FROM for wanderer, never about an update. Its step `v0.20.0 → v0.21.0` (web + db) and meilisearch `v1.36 → v1.54` stay untested; **the brief's question — does the search index survive meilisearch's move or is it rebuilt — is NOT measured.** **Needs:** a bench venue that gives the stack the DB name (an `extra_hosts` + plain-http override in the bench's render only, stated in the verdict), or wanderer proven on the box venue alone with the operator's word. `audits/more-night-apps-2026-09-30/B/wanderer-probe.txt` **-- 2026-09-30 late: the bench CAN run wanderer now** — `upgrade-test.py` `BENCH_ENV_OVERRIDES` points `PUBLIC_POCKETBASE_URL` at the database container on the bench only (the only address the web image reads, measured in v0.20.0 and v0.21.0), recorded in every verdict: web 200, sign-up (`PUT /api/v1/user`) 200, login 200. **The meilisearch question, answered:** v1.54.2 REFUSES v1.36's database („Your database version (1.36.0) is incompatible”); with `MEILI_UPGRADE_DB=true` it upgrades in place and all three indexes (actors, lists, trails) are there — so wanderer's step needs that switch in the template first. Not done: a fixture (creating a list answered PocketBase's „Failed to create record” — likely a rule for unverified users), whether a household's record survives in the index, the step itself. `audits/night-rulings-2026-09-30/` | **NARROWED — the bench runs it; the step needs a fixture and `MEILI_UPGRADE_DB`; owner: CC** **Re-ranked 2026-10-03: P3→P4: test bench coverage; the template switch and fixture are tooling work.** | — | — | CC |
|
||||
| **R-759** | Process & tooling | P4 | **[P3-LOW] wger's onboarding record (the checklist pilot) keeps rows open that no other row owns.** 2026-10-01, `app-catalog-felhom.eu/onboarding/wger.md`: **2.5** no backup → remove → restore → read back of wger exists — the box has no per-app backup press outside an Update (R-648) and wger has no newer step to carry one; **3.7** changing the password and adding a family member not measured, and the template has no `add_people` text; **6.3** no forced-fail undo for wger; **8.2** the app page not read on 9202 this session; **9.1** the runtime volume-persistence gate not re-run (last CLEAN 2026-08-02). The other open rows have their own: 1.5 (R-755), 1.7/2.8 (R-762), 3.4 (R-763), 7.1 (R-764). wger is exempt from the onboarding gate (published before the checklist), so nothing blocks; this row is what keeps the record honest. **Needs:** the five measured on 9202 — 2.5 and 6.3 ride wger's next ladder step (the update's backing-up phase is the per-app backup). `audits/new-app-checklist-2026-10-01/` | **READY — rank P3-LOW; owner: CC (catalog)** **Re-ranked 2026-10-03: P3→P4: record-keeping for a hidden app; nothing blocks.** | — | — | CC |
|
||||
@@ -499,8 +472,6 @@ stopping line that lies.
|
||||
| **R-805** | Process & tooling | P4 | **[P3-LOW] The volume-persistence gate judges an empty NAMED volume (R-788) but not an empty BIND mount.** MEASURED 2026-10-02 (`audits/persistence-sweep-2026-10-02/A/sweep/TABLE.md`): Grimmory's `/app/data` bind held 0 files and read CLEAN (its state is in MariaDB); komga, paperless-ngx, radarr and sonarr also carry empty binds that are not judged. **Needs:** decide whether an empty bind after the exercise is UNDETERMINED like a named volume, with a decoy. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: a test-gate rule; no household meets it.** | — | — | CC |
|
||||
| **R-806** | Process & tooling | P4 | **[P3-LOW] The gate's GET exercise speaks plain http to an HTTPS backend (crafty-controller :8443), and gramps-web did not answer on :5000.** MEASURED 2026-10-02 (`audits/persistence-sweep-2026-10-02/A/sweep/TABLE.md`): both reached data only through their fixture seed. **Needs:** read the `loadbalancer.server.scheme` label (upgrade_boxport already does); find why gramps-web's first GET got no answer. | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: a test-gate defect; no household meets it.** | — | — | CC |
|
||||
| **R-807** | Process & tooling | P4 | **[P3-LOW] 13 apps stay UNDETERMINED because their seeds write only to the database, so their upload / media / cache / redis volumes stay empty.** MEASURED 2026-10-02 (`audits/persistence-sweep-2026-10-02/A/sweep/TABLE.md`): claper, crafty-controller, dawarich (redis), docmost, gramps-web, immich (ML cache), outline (+redis), sparkyfitness, tandoor, vikunja, wger, wishlist, zipline — plus plex (no seed route: a plex.tv claim token) and wanderer (no fixture; unhealthy under the gate). None is BROKEN: nothing was written outside a preserved folder. **Needs:** per app, a seed that uploads one file (or the volume named n/a with a reason: a redis cache, an ML model cache). | **READY — rank P3-LOW; owner: CC** **Re-ranked 2026-10-03: P3→P4: test coverage; nothing was found broken.** | — | — | CC |
|
||||
| **R-817** | Process & tooling | P4 | **`09` decision 56 and R-745 disagree about what the controller self-update rolls back to.** Decision 56 (`documentation/architecture/09-update-architecture.md:619`) says a box keeps the image before the running one as *"the self-update's roll-back target"*; R-745 (closed, `CLOSED-ITEMS.md`) measured that *"the agent rolls back to the RUNNING image (what `/etc/felhom-controller-image` named), never to a previous one the controller hands it"*. Found 2026-10-03 while carrying R-745's rule to `CONTEXT.md`. | **VERIFY — filed 2026-10-03 (triage); owner: CC.** Read the agent's rollback path and correct whichever text is wrong. | — | — | CC |
|
||||
| **R-818** | Process & tooling | P4 | **Two changelogs cite register ids for other findings.** `hub/CHANGELOG.md:526-538` (hub v0.109.0, the Backup card) and its note at `:533` use **R-331** and **R-330** for a hub display fault and a nightly false app alarm; in the register R-330 and R-331 are the disk-health Phase 2 and Phase 3 rows. A reader following the id lands on the wrong finding. Found 2026-10-03 by the triage. | **READY — filed 2026-10-03 (triage); owner: CC.** Add a dated correction line under each changelog entry naming the right rows (the real ids are in `CLOSED-ITEMS.md`); do not renumber anything. | — | — | CC |
|
||||
| **R-819** | Process & tooling | P4 | **`scripts/check_stands.py` is red and runs in no runner.** Measured 2026-10-03 on `9e2786c` (before the triage): it convicts `where-felhom-stands.yaml` for citing R-273 and R-356, which are in neither `OPEN-ITEMS.md` nor anywhere it reads. After the triage it also convicts R-281, R-198 and R-201, because its rule 3 reads only `OPEN-ITEMS.md` and those rows are closed. It is in neither `repo_gates.py` nor CI, so nobody saw it — the R-29 shape. | **READY — filed 2026-10-03 (triage); owner: CC.** Let rule 3 accept an id in `CLOSED-ITEMS.md` (and check the stand's status agrees), fix the two dangling ids, then register it in `repo_gates.py` with a decoy. | — | — | CC |
|
||||
| **R-857** | Process & tooling | P4 | **Baking a golden twice under the SAME version leaves two stale facts.** 2026-10-04 (golden 0.292.0 re-baked with live-restore): (1) `golden_currency_gate.py` keeps reporting the FIRST bake directory's sha (`golden-0.292.0-2026-10-04`, d6cf8b33…) — it picks one of two directories with the same version, not the newest; (2) the publish replaces the package in place (pre-delete + upload), so until the operator re-vouches, the hub vouches a sha the registry no longer holds — a fresh install in that window fails its sha check (fail-closed; none ran; re-vouched 17:48). Fix direction: the gate prefers the newest bake directory (or refuses two for one version); the runbook says "re-vouch at once after a same-version re-bake". `documentation/tests/golden-0.292.0-2026-10-04-rebake/` | **READY — owner: CC** | — | — | CC |
|
||||
|
||||
|
||||
@@ -758,6 +758,8 @@ expensive mistake. That asymmetry is the same one `looksLikeRepositoryDamage` is
|
||||
|
||||
## v0.109.0 — the Backup card told every operator that every customer had no backups (2026-08-30, R-331)
|
||||
|
||||
> **Correction 2026-10-05 (R-818):** the ids **R-330** and **R-331** in this entry are mislabels. These two findings never had register rows of their own; the register's R-330 and R-331 are disk-health Phase 2 and Phase 3. Read „R-330" here as *the nightly false app alarm* (controller v0.224.0) and „R-331" as *the off-site snapshot count read as a measured zero* (hub v0.109.0 Backup card, controller v0.225.0 `stats_known`). Nothing was renumbered.
|
||||
|
||||
> **This push used `git push --no-verify`, and that is declared here rather than worked around.**
|
||||
> `golden_currency_gate.py` was CONVICTED: controller v0.224.0 and v0.225.0 are released and the newest
|
||||
> golden bake carries 0.223.0, so a machine installed right now receives neither. **A bypass, not a
|
||||
|
||||
Reference in new issue
Block a user