diff --git a/CONTEXT.md b/CONTEXT.md index 95981265..aa7485a4 100644 --- a/CONTEXT.md +++ b/CONTEXT.md @@ -16,6 +16,15 @@ > and holds nothing of its own; this file does hold its own content, namely the standing rulings below. +> **2026-11-19/20 (attended) — release 4 is OUT.** Report: `REPORT-release-4-2026-11-19.md`. Rulings: `09` §3 decisions +> 317–319 (D88 = by time; the five own names; the texts as built; the tag with the website). Live: hub 0.183.0, controller +> 0.354.0 (global floor), golden 0.354.0 (vouched), installer 1.38.0 (published). What a session must know: (1) **the test +> household for the off-site end is demo-hp**, not the scratch guest 9202 (no hub link, no off-site tier, the demo-hp id). +> demo-hp's actualbudget is pressed (told 2026-11-19 09:27:34 UTC); do not install actualbudget on demo-hp before +> 2026-12-22, or the proof is voided. The window of 2026-11-24 must refuse it; the one of 2026-12-22 is the first that may +> remove (DUE-CHECKS, R-1374). (2) **There is no per-household switch for the weekly windows** (R-1411). (3) The bind +> form takes extra installer flags (plain `--flag` words only, `store.CleanApplianceExtraArgs`). + > **2026-11-15 evening (unattended) — release 4, second build; nothing is released.** > Report: `REPORT-build-release-4b-2026-11-15.md` (it holds one table for ALL of release 4). Rulings: `09` §3 decisions > 313–316 (D86 remove, D87 keep, the notice lock, the text changes). Heads: `felhom.eu` hub code to `efea8382`, diff --git a/REPORT-release-4-2026-11-19.md b/REPORT-release-4-2026-11-19.md new file mode 100644 index 00000000..e480bef2 --- /dev/null +++ b/REPORT-release-4-2026-11-19.md @@ -0,0 +1,92 @@ +# Release 4 „the promises” — the release run (2026-11-19, attended; ended 2026-11-20 08:45 CET) + +Read first: `documentation/architecture/07-backup-architecture.md` §8.2f, `08-alarm-ladder.md` §12.3, `05-hub-architecture.md` +§14.3, `16-storage-physical-identity.md` §14.5, `runbooks/target-selection.md`, the `felhom-build-deploy` skill, both build +reports of release 4. Evidence: `documentation/audits/release-4-2026-11-19/` and `documentation/tests/golden-0.354.0-2026-11-19/`. + +## 1. Claims in the prompt that were wrong + +| Claim | What is true | +|---|---| +| „The scratch guest can be the test household” (my own evidence file of 11-15) | **Wrong.** Guest 9202's controller has the hub link and the off-site tier OFF and carries the demo-hp household's id. Switching its link on would make it a second reporter of demo-hp (R-1405). The test household was **demo-hp** (guest 9201, Tier 0). | +| „The bind handler's value reaches the installer once the page has the field” | **True as far as it was seen:** the field is live (seen beside a real unclaimed appliance); a test shows the value in the pairing answer; the unit's append is read in source. No bind through the public image was run (a bind is the operator's session act). | +| „A controller on 0.353.0 works unchanged with the new hub” | **True:** three boxes on 0.353.0 reported to hub 0.183.0 for 52 minutes; nothing refused, no error line. | +| „The operator gets about four mails” | **True: exactly 4 notices**, all in the inbox (2 to the original address, 1 to the second address, 1 to the original again). Also one lost-mail summary (see §4). | +| „About 4 hours” | **Wrong.** Active work about 4 h; the run waited overnight at STOP 2's last question and ended 2026-11-20 08:45. | +| „The operator switches the weekly windows off for that household” (`07` §8.2f) | **Wrong:** no such switch. The only switch stops the windows of every household. R-1411. | +| „A screenshot of each” dialog | Not possible here: no browser on DooPlex. The page's script was read (both branches, both languages) and the API's date. | +| „Count one living app's off-site snapshots” | The product shows no per-app count, and a hand `restic` listing is outside the product's flows (forbidden). The store's total (242) is written down; the day-30 read uses the window's own answer. NOT-A-FINDING: the window answer names what it removed. | + +## 2. Six lines + +1. **Hub 0.183.0 is live**, with the five own names (first daily read 2026-11-20 06:40: all five read). +2. **The notice:** seen in the hub's log — yes; in the inbox — yes (4 mails). +3. **The take-back and the second press** — proven on demo-hp (box, hub and inbox), and an address change too. +4. **Every box on controller 0.354.0** — yes (Tester 1, demo-hp, the N100; global floor 10:32 CET). +5. **The installer's three proofs** — 64 GB refusal: yes; the flag's lines: yes (preflight only); 200 GB whole install: yes. +6. **Golden 0.354.0 vouched** — yes (11:32:00 CET). **Installer 1.38.0 tagged and published** — yes (2026-11-20 07:36 UTC, with the website on `main`). + +## 3. The day-30 read + +**2026-12-22** — demo-hp's first weekly window (Tuesday 03:18 UTC) after 30 days by both hub clocks (told 2026-11-19 +09:27:34 UTC). Written in: the DUE-CHECKS block of `OPEN-ITEMS.md` (R-1374), the R-1374 row, `STATUS.md` „do”, `CONTEXT.md`. +Also dated: **2026-11-24**, the window that must refuse the app. Until 2026-12-22 the removal itself is proven by tests only. + +## 4. Facts + +**Versions and digests** + +| Part | Version | Image / file | From | CI (by hash) | +|---|---|---|---|---| +| Hub | 0.183.0 | `sha256:232cf7474bfac3b2ffa3d6fbed105a7359a33d33cb3d5cd6b2341d08d44a9ccd` | `064941f3` | 2710 success; manifest in `bcb628c3`, 2712 success | +| Part 0 field | — | — | `11879001` | 2709 success | +| Controller | 0.354.0 (rc.1 first) | `sha256:c88522ed0ce607f5545b6406276969bb3aa4cc59375e5413bec0537766064e09` (rc.1 `sha256:9cfe81a2…`) | `440711b` | 2713 success | +| Golden | 0.354.0 | `sha256:d680a3e019b8124c67d2cb28c184d4a68ddee6251e8e2193847f37bae6fbbddf` | build-golden.sh `645b3b65…` | — | +| Installer | 1.38.0 | `sha256:4927d98c188a35e365aa7460471900cc90679c353a6c68bbfb312a3a3092051c` | tag `installer-v1.38.0` = `92853bed`; pointer `ea62948c` | 2717 success | +| Agent | 0.175.1 (no change) | `sha256:38b1e235…4612` | — | — | + +**Part 0.** The bind form's field (R-1398): 4 tests, 5 mutations red. The tests against §8.2f: 62 named tests, 0 +missing; every numbered rule has a test (table in the evidence README). The hub backup at STOP 1: the 08:00 copy, on +ep0 at 08:31. A dated check (R-1035, the operator's Search Console read) blocked the push: moved to 2026-11-26 with the +reason in its row. + +**Rulings** (`09` §3 decisions 317–319): D88 = by time; the five names; the texts as built; the tag; publish with the website. + +**Rows:** before 254, after 272. Mine: opened 1 (R-1411), closed 1 (R-1370). The other session opened 18 (R-1406–R-1410, +R-1420–R-1432). Narrowed with dated lines: R-1374, R-1379, R-1376, R-1398. Capability map: C18 added, PARTIAL. +`unproven.py --summary`: no number moved (23 of 55 not walked). + +**Machines changed and how each was left** + +| Machine | Change | Left | +|---|---|---| +| DooPlex | hub 0.183.0 deployed (manifest, hub only); the drill VM booted for the bake | hub Synced/Healthy; drill VM off, reverted to `virgin` | +| demo-hp | guest 9201: rc.1 then 0.354.0; actualbudget installed, backed up off-site twice, removed with its backup data; one end record (2026-12-19); the notification address set on the box (the same one the hub holds); VM 9387 made and destroyed; the ISO downloaded and removed | healthy, 26 containers; rebooted itself 2026-11-20 05:31 for its new kernel (normal path) | +| Tester 1 | per-box floor → 0.354.0 | healthy; floor cleared | +| N100 | the global floor only (no hand step) | 0.354.0, healthy | +| Hub | floors; the test household's notices; host `Tester4-6e517f` and appliance 59 from the proof VM | **operator:** discard appliance 59, delete host Tester4-6e517f | +| Off-site store (demo-hp's) | two „run now” backups (product flow) | 242 snapshots | + +**Proven by tests only:** the removal of an ended app's snapshots in a window (until 2026-12-22); the operator in place +of a household with no address; the stuck line; the daily limit; the hub's „System disk: small” line (no whole install +with the flag); a bind with `extra_args` through the public image. + +**Observations** +- `FILED: R-1411` — no per-household window switch. +- `NOT-A-FINDING:` the lost-mail scan at the hub's start told the operator of 5 operator alarms lost in DooPlex's + internet outage (2026-11-19 02:13–02:55 CET; the mail service timed out) — R-1383's scan doing its job, first real use. +- `NOT-A-FINDING:` the box refuses to empty its address field without „confirm clear” — by design; confirming would + remove the address on the hub too, so the box's field now holds the hub's address. +- `NOT-A-FINDING:` a „run now” off-site backup already asks the hub, so the box learned `accepts_ended` at once, not at + the next night — earlier than the build's note said, and harmless. +- `NOT-A-FINDING:` the proof box's alarms (`backup_tier_skipped` at install; stale and down after demo-hp's reboot left + the VM off) — expected for a throwaway box. +- `NOT-A-FINDING:` a per-box floor for the N100 was refused by Claude Code's permission check; the N100 took the release by + the global floor, as the prompt asked („the N100 by the normal path"). + +**Mistakes of mine** +- The first key-typing helper echoed the keys: the proof VM's throwaway root password showed in my tool output. The + installer replaced it (step 4b) and the VM is destroyed. +- A push waiter hung twice (session 2's uncommitted files blocked `pull --rebase`; once I gave a short hash). Both stopped + by PID; the pushes themselves landed. +- One search of the earlier transcript was refused by the permission check; I did not retry it. diff --git a/STATUS.md b/STATUS.md index 0c2d0c3d..ee533431 100644 --- a/STATUS.md +++ b/STATUS.md @@ -2,7 +2,7 @@ **Ready for Tester 2? Yes — tell the friend go (`documentation/pilot/TESTER2-handover.md`).** -**Updated 2026-11-15 (late evening).** Release 4's second build is in source; nothing is released. New: no notice, no deletion; the delete journal drops the id after a year; the hub shows boxes accepted on a small disk (`REPORT-build-release-4b-2026-11-15.md`). +**Updated 2026-11-20.** Release 4 is out: hub 0.183.0, controller 0.354.0, golden 0.354.0, installer 1.38.0; the deletion notice and its take-back were proven on demo-hp (`REPORT-release-4-2026-11-19.md`). The deletion itself is first seen 2026-12-22. ## What works - **The Plex attack did not reach Felhom's secrets** (R-1241). @@ -34,8 +34,7 @@ Newest first. CC's pick is in the „Pick" column. Each row also has its full te |---|---|---|---|---|---|---| | D91 | R-1410 | **Answer today.** The scrambled form of the hub's FIRST password has been in the public repository since 2026-02-16, and the hub has no limit on wrong tries. Did you change the hub password on the Configuration page after that date — and will you change it now in any case? | **Change it now** (one minute on the Configuration page; then the public value opens nothing while the stored one is readable. CC could not read which password is live) | keep it: if the February password is still the live one, a stranger can guess at it offline with no trace, and it opens everything, the boxes' emergency passwords too — then R-1410 is a P1 | the same as „keep it” | the hub fix task (hash out of git, try limit, sign-out) | | D90 | R-1422 | A taken-over hub can today give any claimed box a new reset code (and so a new dashboard password), and can choose where a box takes its apps from. Close this by design, or accept it for the beta? | **Close it** (a reset of a claimed box, and a change of the box's app source, need your signature, like an agent update; the hub then cannot do either alone) | accept: nothing is built; we write into the threat model that the hub alone can take over dashboards, and the security text for customers must not say otherwise | it stays as it is, and no document says so; the next reader files it again | the design note for R-1422; nothing else | -| D89 | R-1374 | When a household's notification address is changed while an off-site deletion waits, send one mail to the OLD address too? | **Yes** (it closes the last way a stolen login deletes copies unseen: today only the new address is told) | no: nothing new is mailed and no old address is kept; a person who changes the address and keeps it changed gets the notice himself | as built: only the address of the day is told | — | -| D88 | R-1374 | How do we prove the off-site deletion before households can use it: by time, or by hand-dated rows? | **By time** (the scratch box presses on the release day and is the first box whose copies go, 30 days later; nothing is written by hand) | by hand: the removal is seen on the release day, but two rows of the live hub's database are written by hand, and the 30-day clocks are not proven | by time | the release of Part A | +| D89 | R-1374 | Send one mail to the OLD notification address at EVERY change of a household's notification address — not only while an off-site deletion waits? | **Yes** (the reviewer's reading: a person with the login changes the address FIRST and presses after, so a mail only while a deletion waits would never go) | no: nothing new is mailed and no old address is kept; such a person gets the notice himself | as built: only the address of the day is told | — | | D66 | R-1365 | Should a security-watch source that stays down mail again every 7 days, as an open notice does? Today it mails once, then only the Monday summary lists it. | **Yes** (a real outage then cannot be forgotten; the seven pins that can never be read no longer count, so it stays quiet when all is well) | no: one line in a long Monday mail stays the only reminder | a source that stays down mails once; the Monday list still shows it | — | | D38 | R-1235 | Put the newest golden (637 MB) into the nightly release copy on ep0 too? | **Yes** (a new install after a DooPlex loss needs it; it fits under the 3 GB cap) | leave it out: after a loss, no new install until a golden is rebuilt and vouched | the same | — | | D37 | R-1236 | Raise Gitea's memory limit from 1 GiB to 2 GiB? It was killed for memory 4 times in 17 days, once during today's release. | **Yes** (DooPlex has the RAM; a DooPlex change, so yours) | keep 1 GiB: a restart in release hours fails box updates and CI for a minute | the same | — | @@ -43,7 +42,7 @@ Newest first. CC's pick is in the „Pick" column. Each row also has its full te | D7 | R-990, R-989 | Rule on the forum ideas and the selfh.st ideas now, or park them? | **Park them** (P4; nothing waits on them) | Rule now: about one hour of reading | nothing changes; the research gets old | — | ## Waiting on you — do -**2026-11-15 evening.** Publish the website again (Securo's card): `scripts/publish-site-cloudflare.sh`. Read `REPORT-build-release-4b-2026-11-15.md` (all of release 4; the texts). **2026-11-15 midday.** 1. Hosts page: discard host `Tester4-2f0c25`, appliance `ab0b8310…` (else it alarms). 2. Once: read each off-site sub-account's label against the hub's list (R-1375). **2026-11-11.** 2. Hosts page: discard `3fb5fed3…`. 3. Rotate login `kisfenyo`'s Gitea token into the store as `GITEA_GIT_TOKEN` (R-925); the push after `sec put` is in R-1390. 4. Home DNS: remove the `felhom.eu` override (R-1389). 5. Read the dates only you can read (R-1381). 6. Search Console read (R-1035, 11-18). +**2026-11-20 (release 4).** 1. Hosts page: discard appliance 59 and delete host Tester4-6e517f (CC's proof VM; else it alarms). 2. Release 4 ends the order of decision 306: the next step is yours. 3. CC reads demo-hp's window on 11-24 (must refuse) and 12-22 (the deletion; R-1374). **2026-11-15.** Once: read each off-site sub-account's label against the hub's list (R-1375). **2026-11-11.** 2. Hosts page: discard `3fb5fed3…`. 3. Rotate login `kisfenyo`'s Gitea token into the store as `GITEA_GIT_TOKEN` (R-925); the push after `sec put` is in R-1390. 4. Home DNS: remove the `felhom.eu` override (R-1389). 5. Read the dates only you can read (R-1381). 6. Search Console read (R-1035, 11-18). **Still needed? (yes/no)** — answer per line. 1. Look at `roms/roms/` on the N100's drive in the file manager (R-1099). (since 10-16) diff --git a/documentation/architecture/00-capability-map.md b/documentation/architecture/00-capability-map.md index dc5793a8..70ec1ca0 100644 --- a/documentation/architecture/00-capability-map.md +++ b/documentation/architecture/00-capability-map.md @@ -162,6 +162,7 @@ | [C15](capability-map-detail/C.md#c15) | A local backup is bounded by the box's FREE SPACE, not by a partition set at build time — the appliance ships ONE data volume, and a capture that would exhaust it is refused per app rather than allowed to stop the container runtime | golden `build-golden.sh` v3.0.0, agent v0.120.0, controller v0.193.1 | **PROVEN-LIVE** | both halves | `audits/SPIKE-r165-phase0-2026-08-03.md` (2026-08-03) | | [C16](capability-map-detail/C.md#c16) | Soft-quota: usage bar, pre-push enlargement block, customer notification | controller v0.109/134, hub v0.41/55 | **PROVEN-LIVE** | — | `audits/CAMPAIGN-6D-2026-07-15.md` (2026-07-15); `audits/CAMPAIGN-6E-2026-07-15.md` (2026-07-15) | | [C17](capability-map-detail/C.md#c17) | A customer (not the operator) performs a restore via UI alone | all | **MISSING** | as evidence | `audits/DIAG-immich-restore-2026-07-19.md` (2026-07-19); R-3 | +| [C18](capability-map-detail/C.md#c18) | A removed app's off-site copies end 30 days after the hub's own mail to the household, and the household can take it back | hub v0.183.0, controller v0.354.0 | **PARTIAL** | demo-hp (test household) | `audits/release-4-2026-11-19/` (2026-11-19); R-1374 — the removal itself is first seen 2026-12-22 | ## D. Storage & devices diff --git a/documentation/architecture/07-backup-architecture.md b/documentation/architecture/07-backup-architecture.md index 10676b2b..36368d6f 100644 --- a/documentation/architecture/07-backup-architecture.md +++ b/documentation/architecture/07-backup-architecture.md @@ -1835,7 +1835,7 @@ minutes). **[FACT]** 9202 rc.6: an image that exits after 20 s → not started, unit back → success, copy released 3 minutes later. Tester 1 on 0.343.0: a unit restore and a set-aside restore, each „ran stably for 3m0s … the restore copy is released" (`audits/release-2026-10-26/`). -### 8.2f The off-site copies of a removed app end 30 days after „delete backup data” `[FACT — built, not released: controller `f628cb5` and hub `01a106b7` + `a63f5b2b` (after v0.182.0), 2026-11-15; the notice lock and the text changes: hub `42ef25c2`, `9ef8cee1`, `c9d67309` and controller `10df99b`, 2026-11-15 evening; the independent review's six fixes: hub `73a534e7`, `508f936a`, `b391a232`, `cbfeeb62`, `a6f4aa0b`, `efea8382`, 2026-11-15 night; NO box has run it]` (D72 and D84, `09` §3 decisions 292 and 309; the notice lock and the texts, decisions 315 and 316; R-1374 part 1; designed 2026-11-14) +### 8.2f The off-site copies of a removed app end 30 days after „delete backup data” `[FACT — LIVE since 2026-11-19 (hub 0.183.0, controller 0.354.0): the press, the notice, the take-back, a second press and an address change proven on demo-hp; the removal itself proven by tests only until 2026-12-22 (R-1374). Built: controller `f628cb5` and hub `01a106b7` + `a63f5b2b` (after v0.182.0), 2026-11-15; the notice lock and the text changes: hub `42ef25c2`, `9ef8cee1`, `c9d67309` and controller `10df99b`, 2026-11-15 evening; the independent review's six fixes: hub `73a534e7`, `508f936a`, `b391a232`, `cbfeeb62`, `a6f4aa0b`, `efea8382`, 2026-11-15 night; NO box has run it]` (D72 and D84, `09` §3 decisions 292 and 309; the notice lock and the texts, decisions 315 and 316; R-1374 part 1; designed 2026-11-14) **The ruling.** When the household removes an app and presses „delete backup data” (`remove_backups`), that app's off-site copies end 30 days later, at the next clean-up after that day — the number of a restore copy (D62). When the @@ -2069,7 +2069,7 @@ clock, and then name snapshots of other apps under it. What bounds that now, all - **The take-back goes through the box.** A household that reads the mail and presses „keep it after all” is safe when its login was used by somebody else (the mail's case, D84). When the box itself is taken over, the box can show „kept” and go on naming the app. The mail's last line then does not help; the operator does (the mail's - sign-off), by switching the weekly windows off for that household. **Open.** + sign-off), by switching the weekly windows off for that household. **Open — and that switch does not exist: the only switch (`POST /offsite/windows-enabled`) stops the windows of EVERY household (read in source 2026-11-19, R-1411).** - **A household with no address is not told; the operator is.** What happens next depends on the operator. - **A changed address is told again** (review fix 1): the notice counts only for the address the household has now. **NOT closed: a person who holds the login AND keeps the address changed** — or who also reads the diff --git a/documentation/architecture/09-update-architecture.md b/documentation/architecture/09-update-architecture.md index 6e5953aa..e891d923 100644 --- a/documentation/architecture/09-update-architecture.md +++ b/documentation/architecture/09-update-architecture.md @@ -1265,6 +1265,12 @@ Recorded before the work (`audits/space-list-2026-11-10/README.md` §5 held the Hetzner and the mail service are read from source and documents only: no hub database read, no account call. *Operator, 2026-11-10 12:22.* +### 2026-11-19 (operator, release 4, STOP 1 and STOP 2) — D88, the own names, the texts, the vouch, the installer tag + +317. **D88 = by time:** the off-site deletion of a removed app is proven on a real store by waiting — the test household (demo-hp) pressed on the release day and is the first box whose copies go, at its first weekly window after 30 days by both hub clocks (2026-12-22). No row of the live hub's database is written by hand. **Cost of the other option (by hand):** two rows of the live hub's database written by hand, and the 30-day clocks not proven. R-1374. *Operator, 2026-11-19, STOP 1.* +318. **The five own names** the hub reads daily are `hub.felhom.eu`, `files.felhom.eu`, `iso.felhom.eu`, `gitea.dooplex.hu`, `felhom.eu` (`08` §12.3, the manifest's `HUB_OWN_NAMES`); **the texts of release 4 are released as built** (`07` §8.2f, the text table). R-1379, R-1374. *Operator, 2026-11-19, STOP 1.* +319. **Installer 1.38.0 is tagged and published together with the website on `main`** (the Securo card, the count 64, the estimator — the other session's commits of 2026-11-15, not yet live); the Tester 2 laptop's system disk is 200 GB or more. R-1370. *Operator, 2026-11-19/20, STOP 2.* + ### 2026-11-15 (evening, operator, the start line of the second build of release 4) — D86, D87, the notice lock, the text changes The operator's line, word for word: „D86 remove. D87 keep. Notice lock yes. Text changes yes.” diff --git a/documentation/architecture/capability-map-detail/C.md b/documentation/architecture/capability-map-detail/C.md index bb51ea40..2024226f 100644 --- a/documentation/architecture/capability-map-detail/C.md +++ b/documentation/architecture/capability-map-detail/C.md @@ -300,3 +300,13 @@ **Gap / roadmap (as written):** Alpha will produce this; script it into R-3. **2026-07-19:** the C6 evidence attempt ran and found a **product gap instead of evidence** — `audits/DIAG-immich-restore-2026-07-19.md`. A customer-driven UI restore of a DB-indexed app cannot currently succeed (R-43 file-only restore, R-44 stale dump), so this row cannot flip until those close. Row stays MISSING **by finding, not by absence of attempt** — the rehearsal system working, not failing. **2026-07-19: the blocking product gaps are CLOSED in controller v0.148.0** (R-43 + R-44 shipped), so this row is now blocked only on the evidence run itself, not on missing capability. It flips the moment the §9 acceptance produces screenshots + the outcome flash + a snapshot ID. **2026-07-19 round 2 — PARTIAL EVIDENCE ONLY, row NOT flipped** (`audits/DIAG-immich-restore-round2-2026-07-19.md`): a deliberate run from snapshot `49e7cb46` did recover all 11 assets (`status=active`, files resolve), but the operation **reported failure** and left immich reporting schema drift, because the replay aborted against the running app (H4). Photos back ≠ clean acceptance. **2026-07-20: H4 closed in controller v0.153.0 (R-47) on BOTH paths, AND THE EVIDENCE RUN HAPPENED.** *(The "closing in v0.149" wording above was wrong — v0.149.0 was the F3 dashboard fix; R-47 shipped in v0.153.0.)* The C6 drill ran end-to-end **through the UI**: photos deleted, **trash emptied**, the full files+database restore pressed on `/backups/restore`, 40 files placed + 1 DB dump replayed rc-0, 11 assets back, no drift, timeline visually confirmed. The method note below is now DEMONSTRATED, not merely written down. Evidence: `felhom-controller/REPORT.md` 4e. **Residual: the run was performed by the OPERATOR, not by a customer** — for this row literal wording the alpha still owes one genuinely customer-driven pass, but no product gap blocks it. Method note for R-3's script: deleting in an app's own UI usually means *trash*, not deletion, so a drill written that way merges 0 files, flashes success and proves nothing — a real drill must empty the trash **and** verify the app's *content*, not the file count **Lane split → `07-backup-architecture.md` §3**: this row is Lane 1 (customer, unassisted). §8 rows 1–5 are the routes it would exercise --- + + +## C18 — A removed app's off-site copies end 30 days after the hub's own mail to the household, and the household can take it back + +*Map row: [C18](../00-capability-map.md#c18) · added 2026-11-20 (release 4).* + +**Status:** **PARTIAL** — live on every box since 2026-11-19 (hub 0.183.0, controller 0.354.0). **Proven on demo-hp:** the remove dialog's day, the box's end record, the Backups page line and button (both languages), the hub's notice and the mail in the inbox, the take-back on the box and on the hub, a second press, an address change (told again at the new address, and again when it was put back). **Not yet seen:** the removal itself — the first weekly window after the 30 days, 2026-12-22 (R-1374, a DUE-CHECK). Design and rules: `07` §8.2f. + +**Evidence:** `documentation/audits/release-4-2026-11-19/` and `REPORT-release-4-2026-11-19.md`. + diff --git a/documentation/audits/release-4-2026-11-19/README.md b/documentation/audits/release-4-2026-11-19/README.md new file mode 100644 index 00000000..860f25b5 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/README.md @@ -0,0 +1,75 @@ +# Release 4 „the promises” — evidence (2026-11-19, attended; ended 2026-11-20 morning) + +Report: `REPORT-release-4-2026-11-19.md` (repo root). No mail address is in this folder; the household and box ids are +the test boxes' own (demo-hp, tester-1, demo-felhom, Tester4). + +## Part 0 + +**The bind form's field (R-1398).** `11879001`: one text input `extra_args` on the Hosts page's bind form; a value +that is not plain `--flag` words is refused (400, and the store refuses it too). Tests `TestR1398_*` (4). Five +mutations seen red in a private copy of `hub/`: the field removed (the render test), the word check off, the word limit +off, the store's check off (the refusal test), the pairing answer dropping the value (two api tests). CI run 2709. +**Live:** the field shows on the Hosts page beside a real unclaimed appliance (the proof VM's, 2026-11-19 09:42 UTC). + +**The tests against the rules of `07` §8.2f.** Every test the section names exists in source: 62 names (and name +prefixes) read from the section, 0 missing (`hub/` and `controller/`). The numbered rules of „No notice, no deletion” +against the tests: + +| Rule (§8.2f) | Test | +|---|---| +| 1 the hub sends the notice itself; „told” after the send returned | `TestNotice_AReportNamesAnApp_TheHubMailsTheHouseholdOncePerPress`, `TestReview4_ToldIsTheTimeAfterTheSendReturned`, `TestNotice_TheBoxesEventMailsNobody` | +| 2 both 30 days | `TestNotice_AgreeEnded_NoNoticeNoDeletion`, `TestD84_AgreeEnded_TheHubsOwnRecordsDecide` | +| 2a told at an address | `TestReview1_ToldDoesNotSurviveAnAddressChange`, `TestReview1_ToldCountsOnlyAtTheAddressOnRecord`, `TestReview1_Window_ToldAtAnotherAddressIsNotTold`, `TestReview1_BackAndForth_Removed_AndDuringTheSend`, `TestReview1_ASaveThatChangesNothingKeepsTheNotice` | +| 2b a plain app slug | `TestReview3_ALinkShapedNameNeverReachesTheHousehold`, `TestReview3_ANameThatCannotBeToldIsNeverAgreed`, `TestReview3_EveryCatalogAppNameIsANoticeName` | +| 3 where „told” is kept | `TestNotice_ADatabaseWithoutTheNoticeTellsAgain`, `TestReview6_*` | +| 4 take-back, install, new press; the tier off is silence | `TestNotice_ATakeBackOrAnInstallVoidsTheNotice`, `TestNotice_Record_WhatVoidsIt` (its case „the report does not show the tier: its silence is not a take-back”) | +| 5 the daily limit delays, never drops; failed tries count | `TestNotice_TheDailyLimitDelaysAndNeverDrops`, `TestNotice_TheDailyLimitDelaysAndSurvivesARestart`, `TestReview2_ASendThatErrorsAfterItWasAcceptedIsNotMailedEveryHour` | +| 6 no address: the operator in its place, bounded | `TestNotice_NoAddress_TheOperatorIsToldInItsPlace_AndAnAddressLaterRestarts`, `TestNotice_NoAddress_ABoxCannotFloodTheOperator`, `TestNotice_NoAddress_AndTheOperatorsMailFails_NobodyWasTold` | +| 7 a notice that cannot go out | `TestNotice_AFailedSendIsTriedAgainUntilItHasGone`, `TestNotice_TheMailHoldTellsNobody`, `TestNotice_StuckForThreeDays_TheOperatorGetsOneLine`, `TestReview7_ARowToldToTheOperatorNeverGetsTheStuckLine` | +| 8 a window for an app never named in a report | `TestNotice_Window_TheNoticeRecordDecides` | +| 9 a database put back from a copy | `TestNotice_ADatabaseWithoutTheNoticeTellsAgain` | +| 10 an older controller: nothing told, nothing agreed | no test of its own; it is rule 2 with no row (`TestNotice_AgreeEnded_NoNoticeNoDeletion`) — and live: three boxes on 0.353.0 reported to hub 0.183.0 for 52 minutes with nothing refused | + +No rule without a test. **The hub database backup at STOP 1:** the 08:00 CET copy of 2026-11-19, pushed to ep0 08:31. + +**The test household.** The scratch guest 9202 is NOT one: its controller has the hub link and the off-site tier off, +and it carries the id of the demo-hp household — switching its link on would make it a second reporter (R-1405). The +test household is **demo-hp** (Tier 0; its own off-site account; one reporter: 8 reports in 2 hours; its notification +address is on the felhom.eu catch-all, so every test mail reaches the operator's inbox). + +## Part 1 — the hub (`hub-0.183.0-start.txt`, `hub-log-release-4.txt`) + +Hub 0.183.0, image `sha256:232cf747…`, from `064941f3` (CI 2710); manifest `6b8880b3` (with `HUB_OWN_NAMES`), pushed +in `bcb628c3` (CI 2712); synced 09:37 CET, Synced/Healthy. The start line names the five names; the first daily read +(2026-11-20 06:40) read all five. The prune ran at the start and at 04:30 with no error. The Hosts and System pages load. +The hub's own app record held rows for the three test boxes after the first reports (demo-felhom 5, demo-hp 13, +tester-1 3). The start's lost-mail scan told the operator of 5 operator mails lost in DooPlex's internet outage of +2026-11-19 02:13–02:55 CET (the mail service timed out) — the scan of R-1383 working, the first time it found something. + +## Part 2 — the press on demo-hp (all times UTC, 2026-11-19) + +| Step | Seen | +|---|---| +| 08:41 rc.1 on demo-hp's guest 9201 by hand; box walk | 26 containers, 0 unhealthy, 0 ERROR | +| 08:42 actualbudget installed; 08:44 and 08:47 two off-site runs | store 210 → 226 → 242 snapshots | +| the dialog's data before the press | `offsite_end_date` `2026. 12. 19.` (HU), `2026-12-19` (EN); `press-0-*`, `press-1-*` (no browser on DooPlex: the page's script read, both branches) | +| 08:56:51 remove with „delete backup data” | the box's record `ends_at 2026-12-19T08:56:51Z`; the page line and button in both languages (`press-2-*`) | +| 08:56:54 the hub | „the household was told; the hub agrees … no earlier than 2026-12-19”; the box's own event recorded, not mailed; mail 1 in the inbox (subject with the day; the „if this was not you” paragraph) | +| 08:58:40 „keep it after all” | the box's record gone; page „A távoli másolatok megmaradnak.” (`press-3-*`); at the 09:11:54 report the hub's notice table empty | +| 09:12:26 „delete in 30 days” | a new record; 09:26:54 the hub's notice, mail 2 | +| 09:27:12 the address changed to a second catch-all address | the hub: „1 off-site end notice … NOT told any more”; 09:27:16 mail 3 to the new address | +| 09:27:30 the original address back | the same line; 09:27:34 mail 4 to the original address | +| 09:27:51 the box's field emptied again | refused by the box (it needs „confirm clear”, which would remove the address on the hub too); the hub unchanged. The box's field now holds the hub's address — the one difference left | + +4 notices that day; the limit is 5. The hub's notice row after it: told 2026-11-19 09:27:34. demo-hp's windows open on +Tuesday 03:18 UTC, so the window of 2026-11-24 must refuse and the one of 2026-12-22 is the first that may remove +(both DUE-CHECKS on R-1374). **The switch:** there is no per-household switch for the weekly windows — R-1411. + +**Floors:** 0.354.0 image `sha256:c88522ed…` from `440711b` (CI 2713). Per-box floors on tester-1 and demo-hp 10:29 CET +(both on 0.354.0 at 10:30, box walk green); global floor 10:32:18 CET, read back (`min_controller_version` 0.354.0, +MinAgent 0.158.1); the N100 took it at 10:32 by the floor, box walk green; per-box floors cleared. From the global +floor to the next morning: no warning or error event on the three boxes. + +## Part 3 + +Golden and installer: `../../tests/golden-0.354.0-2026-11-19/`. diff --git a/documentation/audits/release-4-2026-11-19/hub-0.183.0-start.txt b/documentation/audits/release-4-2026-11-19/hub-0.183.0-start.txt new file mode 100644 index 00000000..af6dc221 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/hub-0.183.0-start.txt @@ -0,0 +1,9 @@ +2026/11/19 09:37:52 [INFO] Default controller-version floor: 0.120.0 +2026/11/19 09:37:53 [INFO] Pruned 84 old report rows +2026/11/19 09:37:54 [INFO] Pruned 84 old app telemetry rows +2026/11/19 09:37:55 [INFO] Pruned 7 stale app issues +2026/11/19 09:37:55 [INFO] prune: next run at 2026-11-20 04:30 CET (in 18h52m4s) +2026/11/19 09:37:56 [INFO] own-names check: 5 name(s) are read daily at 06:40 Budapest: hub.felhom.eu, files.felhom.eu, iso.felhom.eu, gitea.dooplex.hu, felhom.eu +2026/11/19 09:37:56 [INFO] daily ledger: own-names-check is watched from now (no earlier record) +2026/11/19 09:37:56 [INFO] own-names-check: next run at 2026-11-20 06:40 CET (in 21h2m4s) +2026/11/19 09:37:56 [ERROR] lost operator mail scan: 5 operator mail(s) were never sent — telling the operator in one mail diff --git a/documentation/audits/release-4-2026-11-19/hub-log-release-4.txt b/documentation/audits/release-4-2026-11-19/hub-log-release-4.txt new file mode 100644 index 00000000..5dc1970a --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/hub-log-release-4.txt @@ -0,0 +1,34 @@ +2026/11/19 09:37:56 [INFO] own-names check: 5 name(s) are read daily at 06:40 Budapest: hub.felhom.eu, files.felhom.eu, iso.felhom.eu, gitea.dooplex.hu, felhom.eu +2026/11/19 09:38:08 [INFO] managed floor SERVED for tester-1: floor 0.353.0, agent requirement "0.158.1" from manifest (golden 0.353.0) +2026/11/19 09:38:08 [INFO] managed floor SERVED for demo-felhom: floor 0.353.0, agent requirement "0.158.1" from manifest (golden 0.353.0) +2026/11/19 09:38:08 [INFO] managed floor SERVED for demo-hp: floor 0.353.0, agent requirement "0.158.1" from manifest (golden 0.353.0) +2026/11/19 09:56:51 [INFO] Event from demo-hp: offsite_end_scheduled (info) — actualbudget: az alkalmazás távoli másolatai törlésre vannak kijelölve. Törlésük napja: 2026. 12. 19. Addig a Mentések oldalon a „Mégis megtartom” gombbal visszavonhatod. +2026/11/19 09:56:51 [DEBUG] offsite_end_scheduled from the box of demo-hp is recorded, not mailed — the hub sends this notice itself from the box's report +2026/11/19 09:56:54 [INFO] Notice offsite_end_scheduled sent for demo-hp/actualbudget +2026/11/19 09:56:54 [INFO] [end-notice] demo-hp/actualbudget: the household was told; the hub agrees to remove this app's off-site copies no earlier than 2026-12-19 +2026/11/19 10:12:26 [INFO] Event from demo-hp: offsite_end_scheduled (info) — actualbudget: az alkalmazás távoli másolatai törlésre vannak kijelölve. Törlésük napja: 2026. 12. 19. Addig a Mentések oldalon a „Mégis megtartom” gombbal visszavonhatod. +2026/11/19 10:12:26 [DEBUG] offsite_end_scheduled from the box of demo-hp is recorded, not mailed — the hub sends this notice itself from the box's report +2026/11/19 10:26:54 [INFO] Notice offsite_end_scheduled sent for demo-hp/actualbudget +2026/11/19 10:26:54 [INFO] [end-notice] demo-hp/actualbudget: the household was told; the hub agrees to remove this app's off-site copies no earlier than 2026-12-19 +2026/11/19 10:27:12 [WARN] notification address of demo-hp changed: 1 off-site end notice(s) that were told at the old address are NOT told any more — the household is mailed again at its new address and the 30 days start again +2026/11/19 10:27:12 [INFO] Notification preferences updated for demo-hp: address set=true, household=true, events=[] +2026/11/19 10:27:16 [INFO] Notice offsite_end_scheduled sent for demo-hp/actualbudget +2026/11/19 10:27:16 [INFO] [end-notice] demo-hp/actualbudget: the household was told; the hub agrees to remove this app's off-site copies no earlier than 2026-12-19 +2026/11/19 10:27:30 [WARN] notification address of demo-hp changed: 1 off-site end notice(s) that were told at the old address are NOT told any more — the household is mailed again at its new address and the 30 days start again +2026/11/19 10:27:30 [INFO] Notification preferences updated for demo-hp: address set=true, household=true, events=[] +2026/11/19 10:27:34 [INFO] Notice offsite_end_scheduled sent for demo-hp/actualbudget +2026/11/19 10:27:34 [INFO] [end-notice] demo-hp/actualbudget: the household was told; the hub agrees to remove this app's off-site copies no earlier than 2026-12-19 +2026/11/19 10:29:23 [INFO] managed floor SERVED for tester-1: floor 0.354.0, agent requirement "0.158.1" from declared (golden 0.353.0) +2026/11/19 10:29:24 [INFO] managed floor SERVED for demo-hp: floor 0.354.0, agent requirement "0.158.1" from declared (golden 0.353.0) +2026/11/19 10:29:33 [INFO] Notification preferences updated for demo-hp: address set=true, household=true, events=[] +2026/11/19 10:32:22 [INFO] managed floor SERVED for demo-felhom: floor 0.354.0, agent requirement "0.158.1" from declared (golden 0.353.0) +2026/11/19 11:32:00 [INFO] Artifact manifest set: agent=0.175.1 golden=0.354.0 min_agent="0.158.1" wrapper_sha=false bundle_sha="07ebf86362ed9d44776416bd5c2c4b6b629e44d34595a8120fd8fc3f4d9277ff" +2026/11/19 11:32:30 [INFO] managed floor SERVED for demo-felhom: floor 0.354.0, agent requirement "0.158.1" from manifest (golden 0.354.0) +2026/11/19 11:44:35 [INFO] managed floor SERVED for demo-hp: floor 0.354.0, agent requirement "0.158.1" from manifest (golden 0.354.0) +2026/11/19 11:44:36 [INFO] managed floor SERVED for tester-1: floor 0.354.0, agent requirement "0.158.1" from manifest (golden 0.354.0) +2026/11/20 05:34:37 [INFO] Notification preferences updated for demo-hp: address set=true, household=true, events=[] +2026/11/20 06:40:00 [INFO] own-names check: hub.felhom.eu ends 2027-01-12 (53 day(s) left) +2026/11/20 06:40:00 [INFO] own-names check: files.felhom.eu ends 2026-12-23 (32 day(s) left) +2026/11/20 06:40:00 [INFO] own-names check: iso.felhom.eu ends 2026-12-27 (37 day(s) left) +2026/11/20 06:40:00 [INFO] own-names check: gitea.dooplex.hu ends 2027-01-03 (43 day(s) left) +2026/11/20 06:40:00 [INFO] own-names check: felhom.eu ends 2027-01-28 (69 day(s) left) diff --git a/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-en.txt b/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-en.txt new file mode 100644 index 00000000..2aedb764 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-en.txt @@ -0,0 +1,3 @@ + (backupResp.data.offsite_end_date ? '
' + + 'The local backups are deleted now. The off-site copies stay for 30 more days. They are deleted on: {date}. Until then you can take this back on the Backups page.'.replace('{date}', backupResp.data.offsite_end_date) + '
' : '
' + + 'The nightly restic snapshots cannot be deleted one by one — they expire automatically under the retention rules.' + '
') + diff --git a/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-hu.txt b/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-hu.txt new file mode 100644 index 00000000..f8da6c32 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-0-dialog-branches-hu.txt @@ -0,0 +1,3 @@ + (backupResp.data.offsite_end_date ? '
' + + 'A helyi mentések most törlődnek. A távoli másolatok még 30 napig megmaradnak. Törlésük napja: {date} Addig a Mentések oldalon visszavonhatod.'.replace('{date}', backupResp.data.offsite_end_date) + '
' : '
' + + 'Az éjszakai restic pillanatképek nem törölhetők egyenként — a megőrzési szabályok szerint automatikusan elavulnak.' + '
') + diff --git a/documentation/audits/release-4-2026-11-19/press-1-backup-data-api.json b/documentation/audits/release-4-2026-11-19/press-1-backup-data-api.json new file mode 100644 index 00000000..a1ef6324 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-1-backup-data-api.json @@ -0,0 +1,3 @@ +{"ok":true,"data":{"stack":"actualbudget","backup_paths":[{"path":"/mnt/sys_drive/felhom-data/backups/primary/actualbudget","size_bytes":92045,"size_human":"120K","exists":true}],"has_backups":true,"offsite_end_date":"2026. 12. 19."}} + +HTTP 200 diff --git a/documentation/audits/release-4-2026-11-19/press-2-backups-page-en.txt b/documentation/audits/release-4-2026-11-19/press-2-backups-page-en.txt new file mode 100644 index 00000000..c1f16e54 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-2-backups-page-en.txt @@ -0,0 +1,4 @@ +actualbudget +Marked for deletion. Deleted on: 2026-12-19. +Keep it after all +Not installed — first install it again , then you can bring its data back. diff --git a/documentation/audits/release-4-2026-11-19/press-2-backups-page-hu.txt b/documentation/audits/release-4-2026-11-19/press-2-backups-page-hu.txt new file mode 100644 index 00000000..f65aa1bf --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-2-backups-page-hu.txt @@ -0,0 +1,4 @@ +actualbudget +Törlésre kijelölve. Törlés napja: 2026. 12. 19. +Mégis megtartom +Nincs telepítve — előbb telepítsd újra , utána hozhatod vissza az adatait. diff --git a/documentation/audits/release-4-2026-11-19/press-3-after-take-back.txt b/documentation/audits/release-4-2026-11-19/press-3-after-take-back.txt new file mode 100644 index 00000000..d92ffd70 --- /dev/null +++ b/documentation/audits/release-4-2026-11-19/press-3-after-take-back.txt @@ -0,0 +1,5 @@ +A távoli másolatok megmaradnak. +Visszaállítás +-- +actualbudget +Ennek az alkalmazásnak a távoli másolatai megmaradtak. diff --git a/documentation/backlog/CLOSED-ITEMS.md b/documentation/backlog/CLOSED-ITEMS.md index c595268a..b68dbacc 100644 --- a/documentation/backlog/CLOSED-ITEMS.md +++ b/documentation/backlog/CLOSED-ITEMS.md @@ -26,6 +26,12 @@ --- +## 2026-11-20 — R-1370 closed: the installer stops below 200 GB, released (full text of the row: `git show ea62948c:documentation/backlog/OPEN-ITEMS.md`) + +| Row | What | Closed | Evidence | +|---|---|---|---| +| **R-1370** | **The host's system disk has no space budget: no document says how much the host's own files may take, and the installer's sizing rule covers the guest only.** (P3) | CLOSED 2026-11-20 — RELEASED: installer 1.38.0 tagged `installer-v1.38.0` at `92853bed` and published on the operator's yes (release 4, STOP 2). Three proofs on one test VM on demo-hp: a 64 GB system disk is refused before anything changes (the machine the same before and after); `--accept-small-disk` prints the two lines with the cost and the preflight passes; on the same disk grown to 200 GB a whole install with golden 0.354.0 links to the hub, is claimed and runs one app. The rule: an appliance install needs a 200 GB system disk; a byo install only warns (`16` §14.5, D83). Open elsewhere: R-1398 (a small machine bound without the flag loops). | `documentation/tests/golden-0.354.0-2026-11-19/installer-proofs/` | + ## 2026-11-15 (late) — R-1399 closed: Securo's card on the apps page, the counts at 64 (full text of the row: `git show 2abfdd91:documentation/backlog/OPEN-ITEMS.md`) | Row | What | Closed | Evidence | diff --git a/documentation/backlog/OPEN-ITEMS.md b/documentation/backlog/OPEN-ITEMS.md index 7a380523..893c57b2 100644 --- a/documentation/backlog/OPEN-ITEMS.md +++ b/documentation/backlog/OPEN-ITEMS.md @@ -219,9 +219,10 @@ stopping line that lies. | **R-1263** | Backup & restore | P4 | **Small items from review round 2:** a LOCAL restore's put-back mails the operator „its off-site restore FAILED” (`felhom-controller/controller/cmd/controller/main.go:1514-1518`, shared via `restore_unit.go:517-518`); Remove — even „keep my data” — deletes undeclared volumes labelled with the app's project, which no tier holds (`internal/stacks/delete.go:763-764`, R-1117); the capped copy's `--delete` reads the source in the daemon's mount view, the emptiness check in the controller's (`tier2.go:834` vs `:464`); stale comments — `felhom-agent/internal/localapi/backup_cap.go:15` („keep-last”) and `restore_crash_putback.go:32` (names a test file that does not exist); the qBittorrent password goes to whatever answers the DNS name `qbittorrent` (`upwatch/explain.go:94,108-113`); kept-copy Remove has a check-to-remove gap (`tier2_kept.go:92-111`); the release-copy restore check trusts hashes stored on ep0 (`runbooks/total-loss-of-dooplex.md`); rrsync without `-munge`/`-wo` (`audits/day-2026-10-23/08-release-copy-D33.txt:6`). | OPEN | — | fix in each repo's next attended pass | CC | | **R-1294** | Backup & restore | P3 | **demo-hp's weekly off-site clean-up was refused by its own guard on 2026-10-26 (window 16): the plan would remove 209 of 359 snapshots, above the week's cap (179) — nothing was deleted, and each refused week adds to the next plan.** Seen in the operator mailbox during the 2026-10-26 release run (03:17Z, `offsite_prune_guard_refused`, error; details `count_before 359, count_after 359, outcome guard-refused`). The cap refusal is kept on purpose (R-824, R-867 closed the false ones); the open question is whether a store that once exceeds the cap can ever prune again by itself, or refuses every week while it grows. demo-hp is the Tier-0 box with many hand-made backups (sweeps, probes). Earlier cap refusals: demo-hp 2026-10-03 (window 1), demo-felhom 2026-10-05 (window 3). **2026-11-11:** the End-dates list leans on this row (`documentation/audits/end-dates-list-2026-11-11/TABLE.md` DB05). | **OPEN — the grant is SET (D55 = 260, `09` §3 decision 268)**: 2026-10-26 19:10:31Z through the operator's one-purpose token `cc-d55-window`, read back `max:260`; it is used at demo-hp's next window request (after its nightly off-site run, ~03:15Z). Close when that window prunes; then the operator revokes the token. **Earlier:** WAITING-ON-OPERATOR — 2026-10-26 evening: CC could NOT set the D54 grant — `POST /offsite/window-grant/demo-hp` (`max_remove=220`) with CC's hub token answered **403 „missing scope offsite:write"** (the hub log shows the refusal; nothing granted). The plan also grows: restic's dry-run 2026-10-26 18:56Z still 359 copies / plan 209 / default cap 179 (before tonight's run); a simulation of the keep rule adds ~14 per night (14 of 18 groups back up nightly), so ~223 at the night of 10-27 and ~237 at 10-28 — a cap of 220 would be refused again. STATUS D55: the operator grants it, with the cap. The grant opens the box's NEXT window request, which comes after its next nightly off-site run. Ruled D54 = prune once (`09` §3 decision 266). **Earlier:** WAITING-ON-OPERATOR — 2026-10-26 (read only, `audits/r1294-offsite-cap-2026-10-26/`): the plan is HONEST (restic's own dry-run: keep 150, remove 209; every removal 7–30 days old; every app keeps ≥ 9 or its last copies; the bulk is one test day, 2026-10-16, 152 snapshots). It can NOT recover by itself: the cap is half the store and grows ~49/week while the plan grows ~84/week. Only demo-hp has this shape (demo-felhom and Tester 1 pruned normally 10-26). The pick is STATUS D54. **Earlier:** OPEN **2026-11-10 (space list, `documentation/audits/space-list-2026-11-10/TABLE.md`):** place O1, listed from documents only (the off-site box was not read in this run — second pass). **2026-11-10:** the endings list leans on this row (`documentation/audits/endings-list-2026-11-10/TABLE.md` EC23-7). | — | read the guard's cap rule (controller off-site prune) against demo-hp's snapshot list: is the plan honest (old, superseded) or not; if honest, say how such a store gets back under the cap (an attended one-time prune, or a cap that follows the store) — a design question for the operator if it changes what a box may delete | CC | | **R-1362** | Backup & restore | P3 | **With the internet down, the Backups page takes over 2 minutes (135 s and 145 s measured; 5 s with only the hub down).** Walked 2026-11-03 on Tester 1: every other page answers at once on the home network. The page waits on something off-site inside its handler; a household that lost its line and opens exactly this page sees a hang. `audits/security-2026-11-02/part5/outage-wan.log` | OPEN — filed 2026-11-03 | — | name the wait with a timing, then give it a short deadline and show „the off-site state could not be read" | CC | -| **R-1374** | Backup & restore | P3 | **Four kinds of copy are kept with no end named, and nobody is reminded of them.** Found 2026-11-10 (the endings list, `documentation/audits/endings-list-2026-11-10/TABLE.md` endings EC02, EC23, EC18, EC27; read in source; 9202 has no off-site store, so not measured). (1) **Off-site snapshots of a removed app**: the removal keeps them by ruling (closed R-474), and the weekly clean-up keeps the last snapshots of each app for ever (`internal/backup/offbox_window.go:80-82`), also for an app that is gone. No rule says when they end. (2) **A whole restore from the second drive** moves the old files beside the restored ones (`internal/backup/tier2_whole.go:52`, `:132`, `:345`); nothing removes them later and no page lists them. (3) **Exported app files** stay in the export folder of the drive; the pages can start, list and read an export, none deletes one, and nothing ages it (`internal/web/handler_export.go:30-87`). **Members (answers marked N):** EC02-4, EC23-4, EC18-4, EC27-4. | NARROWED — **2026-11-15 evening (second build): the notice lock and the text changes are BUILT, not released** (hub `main` after 0.182.0, controller `main` after 0.353.0; `07` §8.2f). The hub sends the household's notice itself from its own table `offsite_end_notices`, at every report until the mail service has accepted it; the hub agrees to an app only 30 days after that, beside the 30 days of absence; no usable address → the operator is told in its place; a take-back, an install or a new press voids the notice; a notice that cannot go out blocks the deletion and tells the operator after 3 days. An independent review found one real hole and five smaller defects, all fixed with tests seen red: the notice stayed „told” after the notification address was changed (now: told at an address — a change tells the address of the day again and the 30 days start again). **Open, said in §8.2f:** a person who holds the login and keeps the address changed gets the mail himself (D89 asks: one mail to the OLD address at a change); R-1404; R-1405; a rooted box can ignore the take-back. **Nothing ran on a box, in an inbox or against a real store.** How to prove it: `audits/build-release-4b-2026-11-15/E/proving-the-deletion-on-a-test-store.md` (D88). Earlier: NARROWED — **2026-11-15 evening — RULED: D87 = keep (decision 313); the notice lock = yes (315); the text changes = yes (316).** Fix shape of the lock: the hub mails the household itself, from its own state, when a report first names an app in the box's pending end records, and tries again until its record says „told”; the hub agrees to an app only 30 days after that record, by its own clock, beside the 30 days of absence; no address → the operator is told once and that starts the 30 days; a take-back or an install voids the notice. Earlier: NARROWED — **2026-11-15 (build of release 4): part (1) is BUILT on both sides, not released** (controller `main` after 0.353.0, hub `main` after 0.182.0; `07-backup-architecture.md` §8.2f `[FACT — built, not released]`). The record, the date before the press, the take-back, the hub's agreement by its own clock, the deletion inside the weekly window by id, the audit's explanation, one mail to the household at the press. An independent review found five defects, all fixed with tests seen red (a box could have the hub agree to the whole store; a take-back or an install during the window still deleted; an import kept the record; a box could word and flood the household's mail; a version skew stalled the clean-up). **Decided by CC unattended (D87 asks the operator):** one window removes at most the cap's size of ended-app snapshots beside the cap, and a request for more than half of the store waits one week with a mail to the operator. **Open limits, said in §8.2f:** a box that stops listing a living app looks like an honest removal (bounded by the cap per week, and by the 30 days); an install during the forget itself loses the copies; a box that names agreed ids but removes others is caught at 07:10, not at the close. **Nothing ran on a box or against a real store.** Parts (2) and (3) are not touched. Earlier: OPEN — **2026-11-15 12:49 — RULED (D84 = yes; `09` §3 decision 309).** Fix shape for part (1): the five parts of `07` §8.2f, with the ended apps' own allowance outside the weekly cap (only for ids the hub agreed to) and ONE mail to the household at the press of „delete backup data” (the app, the end date, how to take it back). Earlier: OPEN — **2026-11-14 (night build):** the design for part (1) is written, not built: `07-backup-architecture.md` §8.2f (the record, the date shown before the press, the take-back button, the deletion inside the weekly window after the hub agrees, 14 edge cases, the texts in Hungarian and English as a proposal). One open decision is in it: do the ended apps' snapshots count toward the window's weekly cap (CC's pick: no). Parts (2) and (3) are not touched. OPEN — filed 2026-11-10 (endings list) **2026-11-10 15:19 — RULED (D72 = yes, by the household's own choice; `09` §3 decision 292).** Not built. | — | Controller (D72): when the household pressed „delete backup data” for an app, its off-site copies end 30 days later, at the next clean-up after that day (the number of a restore copy, D62); when the household kept the backups, the off-site copies stay. The page shows the end date before anything is deleted — write that into `07-backup-architecture.md` before the build. Parts (2) and (3): a delete on the export page; a line on the kept-data page for the files a whole restore set aside. | CC | +| **R-1374** | Backup & restore | P3 | **Four kinds of copy are kept with no end named, and nobody is reminded of them.** Found 2026-11-10 (the endings list, `documentation/audits/endings-list-2026-11-10/TABLE.md` endings EC02, EC23, EC18, EC27; read in source; 9202 has no off-site store, so not measured). (1) **Off-site snapshots of a removed app**: the removal keeps them by ruling (closed R-474), and the weekly clean-up keeps the last snapshots of each app for ever (`internal/backup/offbox_window.go:80-82`), also for an app that is gone. No rule says when they end. (2) **A whole restore from the second drive** moves the old files beside the restored ones (`internal/backup/tier2_whole.go:52`, `:132`, `:345`); nothing removes them later and no page lists them. (3) **Exported app files** stay in the export folder of the drive; the pages can start, list and read an export, none deletes one, and nothing ages it (`internal/web/handler_export.go:30-87`). **Members (answers marked N):** EC02-4, EC23-4, EC18-4, EC27-4. | NARROWED — **2026-11-19 (release 4): LIVE** — hub 0.183.0 (deployed 2026-11-19 09:37 CET) and controller 0.354.0 (global floor 10:32 CET). **Proven on the test household demo-hp** (not the scratch guest: 9202 has no hub link and no off-site tier, and switching its link on would make it a second reporter of demo-hp, R-1405): a throwaway app (actualbudget), two off-site runs, then „delete backup data” → the dialog's day (`2026. 12. 19.` / `2026-12-19`; in that branch the older sentence is not shown), the box's record, the Backups page line and button in both languages; at the report the hub's notice and the mail in the inbox (the subject with the day, the „if this was not you” paragraph); „keep it after all” → the box's record gone, at the next report the hub's row gone, the page „kept”; „delete in 30 days” → a new notice; an address change → the notice un-told and mailed to the new address, then the old address back and mailed again. 4 notices in all, under the limit of 5 a day. The page's day and the mail's day are the same. **The removal itself: tests only until 2026-12-22** — the first weekly window after the 30 days by both hub clocks (told 2026-11-19 09:27:34 UTC; demo-hp's windows open on Tuesday 03:18 UTC); the read is a DUE-CHECK. The window of 2026-11-24 must refuse the app (DUE-CHECK). Before: 242 off-site snapshots in demo-hp's store (the product shows no per-app count; the day-30 read compares the window's own answer). **Found:** the operator has no switch for ONE household's weekly windows (R-1411). Evidence `documentation/audits/release-4-2026-11-19/`. Earlier: NARROWED — **2026-11-15 evening (second build): the notice lock and the text changes are BUILT, not released** (hub `main` after 0.182.0, controller `main` after 0.353.0; `07` §8.2f). The hub sends the household's notice itself from its own table `offsite_end_notices`, at every report until the mail service has accepted it; the hub agrees to an app only 30 days after that, beside the 30 days of absence; no usable address → the operator is told in its place; a take-back, an install or a new press voids the notice; a notice that cannot go out blocks the deletion and tells the operator after 3 days. An independent review found one real hole and five smaller defects, all fixed with tests seen red: the notice stayed „told” after the notification address was changed (now: told at an address — a change tells the address of the day again and the 30 days start again). **Open, said in §8.2f:** a person who holds the login and keeps the address changed gets the mail himself (D89 asks: one mail to the OLD address at a change); R-1404; R-1405; a rooted box can ignore the take-back. **Nothing ran on a box, in an inbox or against a real store.** How to prove it: `audits/build-release-4b-2026-11-15/E/proving-the-deletion-on-a-test-store.md` (D88). Earlier: NARROWED — **2026-11-15 evening — RULED: D87 = keep (decision 313); the notice lock = yes (315); the text changes = yes (316).** Fix shape of the lock: the hub mails the household itself, from its own state, when a report first names an app in the box's pending end records, and tries again until its record says „told”; the hub agrees to an app only 30 days after that record, by its own clock, beside the 30 days of absence; no address → the operator is told once and that starts the 30 days; a take-back or an install voids the notice. Earlier: NARROWED — **2026-11-15 (build of release 4): part (1) is BUILT on both sides, not released** (controller `main` after 0.353.0, hub `main` after 0.182.0; `07-backup-architecture.md` §8.2f `[FACT — built, not released]`). The record, the date before the press, the take-back, the hub's agreement by its own clock, the deletion inside the weekly window by id, the audit's explanation, one mail to the household at the press. An independent review found five defects, all fixed with tests seen red (a box could have the hub agree to the whole store; a take-back or an install during the window still deleted; an import kept the record; a box could word and flood the household's mail; a version skew stalled the clean-up). **Decided by CC unattended (D87 asks the operator):** one window removes at most the cap's size of ended-app snapshots beside the cap, and a request for more than half of the store waits one week with a mail to the operator. **Open limits, said in §8.2f:** a box that stops listing a living app looks like an honest removal (bounded by the cap per week, and by the 30 days); an install during the forget itself loses the copies; a box that names agreed ids but removes others is caught at 07:10, not at the close. **Nothing ran on a box or against a real store.** Parts (2) and (3) are not touched. Earlier: OPEN — **2026-11-15 12:49 — RULED (D84 = yes; `09` §3 decision 309).** Fix shape for part (1): the five parts of `07` §8.2f, with the ended apps' own allowance outside the weekly cap (only for ids the hub agreed to) and ONE mail to the household at the press of „delete backup data” (the app, the end date, how to take it back). Earlier: OPEN — **2026-11-14 (night build):** the design for part (1) is written, not built: `07-backup-architecture.md` §8.2f (the record, the date shown before the press, the take-back button, the deletion inside the weekly window after the hub agrees, 14 edge cases, the texts in Hungarian and English as a proposal). One open decision is in it: do the ended apps' snapshots count toward the window's weekly cap (CC's pick: no). Parts (2) and (3) are not touched. OPEN — filed 2026-11-10 (endings list) **2026-11-10 15:19 — RULED (D72 = yes, by the household's own choice; `09` §3 decision 292).** Not built. | — | Controller (D72): when the household pressed „delete backup data” for an app, its off-site copies end 30 days later, at the next clean-up after that day (the number of a restore copy, D62); when the household kept the backups, the off-site copies stay. The page shows the end date before anything is deleted — write that into `07-backup-architecture.md` before the build. Parts (2) and (3): a delete on the export page; a line on the kept-data page for the files a whole restore set aside. | CC | | **R-1404** | Backup & restore | P3 | **A household that presses „delete backup data” while its off-site tier is switched off gets no notice until the tier is on again.** Found 2026-11-15 (the notice lock, `07` §8.2f, `[DESIGN — proposed, not built]`). The controller names its pending end records in its report only while the off-site tier is ON (`backup/offbox.go`, `OffboxStatusForReport`), so the hub does not learn of the press and mails nobody. Nothing is deleted meanwhile (the safe direction); when the tier is on again the notice goes and the 30 days start then. The household sees a date on its page that the hub does not keep. | OPEN — filed 2026-11-15 | — | the controller reports pending records with the tier off too — a change of the report's shape, so both sides and the wire gate | CC | | **R-1405** | Backup & restore | P4 | **Two controllers that report for one household undo each other's off-site end notice.** Found 2026-11-15 by the review of the notice lock. A controller's report is taken on the household's one key and stored per household; one household is expected to have one box, and nothing prevents two (a guest that was cloned or restored and kept its config). The notice is then made by one report and voided by the next: up to 5 mails a day to the household, and the deletion it asked for never comes. Safe for the copies. | OPEN — filed 2026-11-15 | — | first: tell the operator when two different controllers report for one household inside one report cycle | CC | +| **R-1411** | Backup & restore | P3 | **The operator cannot stop the weekly off-site clean-up of ONE household; the only switch stops it for every household.** Found 2026-11-19 (release 4, Part 2). `07` §8.2f („What a taken-over box can still do”) and the notice mail's sign-off rely on the operator „switching the weekly windows off for that household”. In source the switch is `POST /offsite/windows-enabled` → `store.SetOffsiteWindowsEnabled`, ONE hub setting (`offsite_prune_windows_enabled`) for all households (`hub/internal/web/offsite_window.go`, `hub/internal/offsitekeys/service.go`). So against a taken-over box that keeps naming an app, the operator's only lever stops the clean-up of every household (old snapshots are then kept everywhere and the stores grow). Not tried: it acts on every household. | OPEN — filed 2026-11-19 | — | a per-household switch on the customer page's clean-up card (no window, or no ended app agreed, for that household), with a test that the global switch still works; then correct §8.2f | CC | | **R-1401** | Backup & restore | P4 | **The controller's fixture tool rewrites unrelated files.** Found 2026-11-15: `-update-i18n-golden` writes `?v=ASSET` where 20 stored page fixtures hold `?v=test`, so a regeneration changes files that the change did not touch; the D72 build patched its 143 fixtures by hand, one line each. Cost: the next change of a shared page string does the same by hand, or commits noise. | OPEN — filed 2026-11-15 | — | make the tool and the stored fixtures agree on the asset version word, then regenerate once | CC | | **R-1385** | Backup & restore | P3 | **Two timed jobs can meet with no lock and no order (12 answers).** Found 2026-11-11 (the Clock list, class W3, `documentation/audits/clock-list-2026-11-11/TABLE.md`). MEASURED on the night of 2026-11-11: on Tester 1 and the N100 the whole-guest backup started at about 05:30 local, the minute of the off-site proof restore (CC09-5); on demo-hp one whole-guest backup started at 02:30:38, with the dump leg; on the N100 the whole-guest backup starts about 5 minutes later each day. From source: the controller's auto-update default is the minute the whole-guest gate opens (CC11-5); the second-drive copy and the abandon sweep wait only for the clock (CC02-5, CC19-5, CC21-5); the PBS verify can run beside a backup (CA13-5, CA04-5); the timers on DooPlex and ep0 are not ordered against the hub's copies (CH05-8, CH24-5, CH25-3, CH26-5, CH26-6; from documents only, not measured). No harm was seen; the durations are from small test stores. | OPEN — filed 2026-11-11 (clock list) | — | Measure one night with a large store on Tier 0 (as a task), then decide per pair: a shared flag, a fixed offset, or „harmless, written down”. | CC | @@ -299,8 +300,7 @@ stopping line that lies. | **R-1367** | Box system & updates | P3 | **The host journal fills a small system disk: 61% of its lines are `pam_unix(sudo:session)` open/close pairs from the agent's sudo calls.** MEASURED 2026-11-09, one hour, same on demo-hp and the N100: ~14,000 journal lines, ~8,500 of them sudo session lines; the top commands are `smartctl -a` (1,074 per hour — one every ~3 s), `lxc-info -n` (662), `mkdir -p` (540), `lvs` (537), `pct exec` (~520). The journal grows by journald's default cap (10% of the filesystem, max 4 GB): the N100 sits at 4 GB, demo-hp was 3.8 GB on 2026-11-07 and reached 0.7 GB again two days after a vacuum to 500 MB. On a 39 GB root that space is what the whole-guest local copy needs (R-548). | OPEN — filed 2026-11-09 **2026-11-10 (space list, `documentation/audits/space-list-2026-11-10/TABLE.md`):** place H7: the N100 (4.28 GB, 74 files) and Tester 1 (4.29 GB, 81 files) sit at the journal's own cap; Tester 1 writes about 150 MB a day. Lines written by `sudo` are 92 / 92 / 90 % of all lines in one hour — this counts three lines for each call (the command and the session pair), while this row's 61 % counts the session pair only: the same fact, counted two ways. No rank change is proposed: the journal stops at its own cap (about 4.3 GB), so it costs space and log history, not safety. | — | Agent, through the signed config bundle: `Defaults:felhom-agent !pam_session` (and `!syslog` only if the command log has no reader — check the audit trail's needs first) in the sudoers file; a journald cap drop-in (e.g. `SystemMaxUse=500M`) on the host; and read why `smartctl -a` runs every ~3 s (the SMART poll's interval or a loop). Red-proof each in the agent's tests; measure the line rate after on demo-hp. Evidence: `/mnt/5_hdd/felhom.eu/evidence/demo-hp-images-2026-11-09/journal-noise.txt` (off-git) | CC | | **R-1368** | Box system & updates | P3 | **Five places on the host's system disk only grow: nothing deletes them and no rule says how many to keep.** MEASURED 2026-11-10 on demo-hp, demo-felhom and Tester 1 (the space list, `documentation/audits/space-list-2026-11-10/TABLE.md` places H5, H6, H9, H12, H13). (1) **Old kernels** (`/lib/modules`, `/boot`): 8 / 8 / 3 kernels = 8.15 / 8.15 / 3.05 GB. The ring-0 boxes got 7 kernels in 37 days and no kernel was ever removed: the update wrapper refuses a plan that removes a package (`11-os-updates.md` rule R4), and nothing runs `apt autoremove` (searched: architecture folder, both registers, runbooks, three source trees — 0 hits). The N100's system disk grew 4.6 GiB in 26 days, about a kernel a week. (2) **The agent's self-update folder** `/var/lib/felhom-agent/selfupdate`: 49 / 48 / 33 binaries = 740 / 724 / 509 MB, one for each agent version the box ever got; no document names the folder. (3) **Guest template archives** `vzdump-lxc-9100-*` in `/var/lib/vz/dump`: 1 / 4 / 1 = 0.65 / 2.56 / 0.65 GB. (4) **Old agent binary copies** `/usr/local/bin/felhom-agent.bak-*`: 1 / 11 / 0 files, up to 155 MB. (5) **Logs of failed whole-guest backup runs** in the dump folder (small; 9 on demo-hp). **Members (answers marked N):** H5-3, H5-4, H6-3, H9-3, H9-4, H9-8, H12-3, H12-4, H13-3, H13-4. | NARROWED — **2026-11-15 06:35 — the FIRST real night removal: the N100, by itself, at 05:58 CET.** 8 kernels before, 2 after (it kept 7.0.14-26 and 7.0.14-28, the running one), 6 packages purged in 35 s, 6.81 GB gained (the wrapper's own figure; 65.1 GB free after), 0 left-over paths; the default kernel and the pinned file are unchanged (the file's time is still 2026-11-08); the hub's „Kernel step” cell reads „cleaned”; 0 failed units, 9 of 9 containers healthy, clock +0.06 s to DooPlex. demo-hp had a QUIET night (2 installed, nothing to remove) — it does not count; its 6 old kernels went on 2026-11-10 at 13:51, by day. **D82 count: the N100 1 of 2, demo-hp 0 of 2.** demo-hp can count only after two more kernels arrive (it holds 2 and keeps 2). Evidence: `audits/night-build-2026-11-14/E/`. Earlier: NARROWED — **2026-11-14 16:06 — RULED (D82, `09` §3 decision 307): a household's box may remove its old kernels, after two REAL night removals on each ring-0 box (the N100 and demo-hp); quiet nights do not count. Count so far: the N100 0, demo-hp 0 (the Tester 1 removal of 2026-11-14 was started by hand and is not a night). It needs a later agent release to switch it on.** Earlier: NARROWED — **2026-11-14 (release 2a): the old-kernel part is BUILT — agent 0.174.0, vouched 15:21 CET (`11` §5.11.1 `[FACT — built]`).** A ring-0 box removes its old kernels at night by the wrapper's own rule. On boxes: a real removal on Tester 1 (3 → 2 kernels, 1.12 GB, 8.5 s; the default, the pinned file, the generators and the state file unchanged; a second run removed nothing; one clean restart) — started by hand with a ring-0 mark set for the test on the operator's yes, because a night cannot be started by day (R-1394); dry runs on demo-hp (nothing to remove; refusals P1, P2, P3) and the N100 (6 of 8 will go). **LEFT for this part: the step by the agent at night — read the N100's and demo-hp's night of 2026-11-15; then a ruling for ring-1 boxes (D82).** Evidence `audits/release-2a-2026-11-14/`. The other four places of this row are unchanged. Earlier: OPEN — filed 2026-11-10 (space list) **2026-11-10 12:22 — RULED (D67 = yes, as a separate step; `09` §3 decision 286):** the box may remove old kernels by itself; rule R4 stays as it is for update plans; the clean-up is its own step with its own narrow rule (old kernel packages only, only after a good boot on the newest kernel). **2026-11-10:** the endings list leans on this row (`documentation/audits/endings-list-2026-11-10/TABLE.md` EA01-4, EA23-4, EA24-4, EA26-4). **2026-11-10 — the kernel part's fix shape is PROVEN by hand on demo-hp** (`documentation/audits/kernel-cleanup-spike-2026-11-10/`; design `11-os-updates.md` §5.11.1, not built): the rule removed 6 of 8 kernels and nothing else, 6.4 GB gained, boot settings unchanged, one orderly reboot back on the same kernel, a second run removed nothing. Not proven: a box whose running kernel is not its newest; a boot of the kept old kernel. **2026-11-10 — part (2) FIXED IN SOURCE, NOT RELEASED** (felhom-agent `776025c8`): after an update that ended well the agent removes every staged binary but the running version's; nothing is removed while an update is pending. `TestPrune_*` (6 tests, one runs the wrapper's real commit and rollback verbs), red seen three times. A box cleans its folder at the first good update after the release that carries it. Part (4), the `.bak-*` copies, is not reached (root-owned, made by a person). Parts (1), (3), (5) stay open. **2026-11-10 14:46 UTC — the kept old kernel boots:** on demo-hp the one-shot flag booted 7.0.14-26-pve to a healthy host in 313 s, and an orderly restart brought 7.0.14-28-pve back in 324 s; the pinned default, `grub.cfg` and the step's state file did not change, and the hub got no kernel report and no alarm (`documentation/audits/kernel-cleanup-spike-2026-11-10/evidence/fallback-boot/`). Still open for the kernel part: a second box; the step built as a program. **2026-11-10 (release run) — part (2) RELEASED, VOUCHED and PROVEN on three test boxes** (agent 0.173.0, vouched by the operator 15:29 UTC; `documentation/audits/release-space-fixes-2026-11-10/agent/`): the folder went 49 → 1, 48 → 1 and 33 → 1 files (about 700 / 690 / 485 MB back) at the commit of THIS update, not at the next one (the earlier sentence here and in the agent's CHANGELOG was wrong, and is corrected there). A rollback after the clean-up worked on Tester 1: a signed update to the older version was crashed once, 9 seconds into the 60-second dwell, and the wrapper put the released version back in the same second; the failed version's file stays until the next good update (by design). The kept older kernel on demo-hp booted to healthy in 313 s and the box went back (`documentation/audits/kernel-cleanup-spike-2026-11-10/evidence/fallback-boot/`). **Still open:** (1) the kernel clean-up as a program (`11-os-updates.md` §5.11.1, not built) and a second box; (3) template archives; (4) the `.bak-*` copies — the release does not reach them; (5) the failed-run logs. A household box gets part (2) at its first good agent update after the vouch. | the kernel part: the hand proof on demo-hp (`documentation/audits/kernel-cleanup-spike-2026-11-10/`), then an agent release | Agent: after a good update keep the running kernel, the one before it and the newest, and remove the rest — ring 0 first; keep the newest N self-update binaries and `.bak` copies; keep one template archive. Each with a test that a kept item is never removed. | CC | | **R-1369** | Box system & updates | P3 | **When a disk is full there is no written way out: no runbook and no self-repair, for the host system disk, the thin pool, the guest disks or a data drive.** Found 2026-11-10 (the space list, `documentation/audits/space-list-2026-11-10/TABLE.md`, question 7). One missing runbook, seen from 21 places; the search of the runbooks folder for a disk-full procedure gave 0 hits (`evidence/searches.txt` S6). The alarms exist (hub: 90 / 95 % host disk, 85 / 90 % thin pool; controller: 85 / 95 % on the data volume and each drive) but they end in a mail, not in a step. For three small places nobody wrote what fails when they are full (EFI partition, host and guest temp folders). **Members (answers marked N):** H1-7, H2-7, H5-7, H9-7, H10-5, H10-7, H12-7, H13-7, H15-5, G1-7, G2-7, G7-7, G12-7, G13-7, G14-7, G17-5, D1-7, D2-7, D3-7, D6-7, D7-7, D8-7, D9-7, D10-7. | OPEN — filed 2026-11-10 (space list) **2026-11-10 (afternoon) — the runbook is written:** `documentation/runbooks/disk-full.md`, five parts, each hand step marked PROVEN (date and evidence; all on demo-hp) or NOT PROVEN. Still open: one rehearsal of the whole runbook on demo-hp; the NOT PROVEN steps (copies on a former target, template archives, the `.bak-*` binaries, the guest root disk); what fails when the EFI partition or a temp folder is full. | — | Docs: one runbook „a disk is full” with one section per disk (host system disk, thin pool, guest root, data volume, data drive): what to look at first, what is safe to delete (the places of R-1368 and R-1367), what is never deleted by hand (household data, the kept-data rulings). Rehearse it once on demo-hp. | CC | -| **R-1370** | Box system & updates | P3 | **The host's system disk has no space budget: no document says how much the host's own files may take, and the installer's sizing rule covers the guest only.** Found 2026-11-10 (the space list, `documentation/audits/space-list-2026-11-10/TABLE.md` place H1, answer H1-3; search S5 — 0 hits). Today: demo-hp 58 % of 41 GB, N100 38 % of 101 GB, Tester 1 34 % of 63 GB — and the parts that grow are the journal (R-1367), old kernels and agent binaries (R-1368), old backup copies (R-1087) and the local whole-guest copy (R-548). Second half, known but with no row until now: on a small disk the guest's own guards read the guest's 70 GiB while the thin pool behind it holds 54 GiB (demo-hp: 102 GiB promised on a 53.9 GiB pool, 70.67 % used), so only the hub's thin-pool alarm stands before a full pool. **Member:** H1-3. | NARROWED — **2026-11-15 (build of release 4): BUILT, not released** — installer v1.38.0 on `main`, NOT tagged (`16` §14.5 `[FACT — built, not released]`): `check_system_disk` first in the preflight; an appliance below 200 × 10⁹ bytes stops; `--accept-small-disk` goes on, prints the cost, records `small_disk`; a BYO host is warned only; an unreadable size stops and names the flag. 8 tests, 15 mutations seen red. **Not run on a machine.** What it does not do is R-1398. For the release: the proof VM gets a 200G disk; demo-hp's reinstall needs the flag. Earlier: OPEN — **2026-11-15 12:49 — RULED (D83 = refuse, with one named flag; `09` §3 decision 308).** Fix shape: the installer reads the size of the disk that holds the `pve` volume group before any change; below 200 × 10⁹ bytes an appliance install stops with the designed sentence; `--accept-small-disk` goes on, prints the cost, records `small_disk: true`; a BYO host is warned only; a size that cannot be read refuses on an appliance and names the flag (`16` §14.5). Earlier: OPEN — **2026-11-14 (night build):** the open point of D75 has a written design with a pick, not built: `16-storage-physical-identity.md` §14.5 — refuse below 200 GB on an appliance install, with one named flag for the operator; warn only on a BYO host; 8 edge cases. It waits for the operator's yes. OPEN — filed 2026-11-10 (space list) **2026-11-10 12:22 — RULED (D69 = yes; `09` §3 decision 288):** the section goes inside `16-storage-physical-identity.md`, no new document. **2026-11-10 (afternoon) — the section is written:** `16-storage-physical-identity.md` §14 (45 places, the list „no limit, by ruling”, the budget, the open design points); `PROMPT-TEMPLATE.md` §N.5 keeps it true. The budget from measured numbers: the root needs about 14 GB when the whole-box copies go to a backup drive, and 28–54 GB at keep 3 when they do not. **D75 (for the operator):** the smallest system disk Felhom accepts. CC's pick: 128 GB only with a backup drive, 200 GB without one. The other option, 128 GB always, lets a box with many apps and no backup drive stop its local copies. With no ruling the installer accepts any disk. Open after the ruling: the installer's refusal. **2026-11-10 15:19 — RULED (D75 = one number, 200 GB, with or without a backup drive; `09` §3 decision 295).** This replaces CC's two-step pick. **Open, not ruled:** what the installer does below 200 GB — refuse, or warn. demo-hp (128 GB) stays as a test box. `16` §14.3 carries the ruling. | — | Design first: what the installer does below 200 GB (refuse, or warn) — a later session. Then the installer. | CC | -| **R-1398** | Box system & updates | P3 | **A machine that the installer refuses for a small system disk tells nobody but its own screen, and the hub does not read the „small disk” mark.** Found 2026-11-15 while building D83 (installer v1.38.0, not tagged). (1) On the public image the first-boot unit (`scripts/iso/felhom-bootstrap.sh`, „retry until success”) runs the installer again every 30 seconds: a refused 128 GB machine prints the refusal on its screen for ever, and the hub shows a bound appliance that never installs, with no reason. (2) An install that went on with `--accept-small-disk` sends `"small_disk":true` in its enrolment request; the hub ignores the field, so no page shows which boxes were accepted small (`16-storage-physical-identity.md` §14.5, the last row of the built table). | OPEN — **2026-11-15 evening (second build): part (2) is BUILT, not released** (hub `main` after 0.182.0): the hub keeps `small_disk` (and a size, when an installer sends one) from the enrolment and the box's page shows „System disk: small — accepted below 200 GB”. **Part (1) is open, and one more thing was read:** the hub's bind form has NO field for extra installer arguments (the handler takes `extra_args`, the page does not offer it; the household's own bind sends none), so `--accept-small-disk` cannot reach an install from the public image today (`audits/build-release-4b-2026-11-15/E/small-disk-flag-on-the-public-image.md`). Earlier: OPEN — filed 2026-11-15 | the next installer ISO for (1) | (1) a final exit code for a refusal that no retry can cure; the first-boot unit stops retrying on it, shows the sentence once, and tells the hub the reason; (2) the hub stores `small_disk` at the enrolment and shows it on the box's page | CC | +| **R-1398** | Box system & updates | P3 | **A machine that the installer refuses for a small system disk tells nobody but its own screen, and the hub does not read the „small disk” mark.** Found 2026-11-15 while building D83 (installer v1.38.0, not tagged). (1) On the public image the first-boot unit (`scripts/iso/felhom-bootstrap.sh`, „retry until success”) runs the installer again every 30 seconds: a refused 128 GB machine prints the refusal on its screen for ever, and the hub shows a bound appliance that never installs, with no reason. (2) An install that went on with `--accept-small-disk` sends `"small_disk":true` in its enrolment request; the hub ignores the field, so no page shows which boxes were accepted small (`16-storage-physical-identity.md` §14.5, the last row of the built table). | OPEN — **2026-11-19 (release 4): part (2) is LIVE (hub 0.183.0), and the bind form's field for extra installer flags is LIVE** (built in Part 0, `11879001`; seen on the Hosts page beside a real unclaimed appliance; only plain `--flag` words pass). Not seen live: the box page's „System disk: small” line (no whole install with the flag was made; the flag's two lines were proven with `--preflight-only`). **Part (1) is open:** a small machine bound WITHOUT the flag repeats its refusal every 30 seconds and tells nobody else. Earlier: OPEN — **2026-11-15 evening (second build): part (2) is BUILT, not released** (hub `main` after 0.182.0): the hub keeps `small_disk` (and a size, when an installer sends one) from the enrolment and the box's page shows „System disk: small — accepted below 200 GB”. **Part (1) is open, and one more thing was read:** the hub's bind form has NO field for extra installer arguments (the handler takes `extra_args`, the page does not offer it; the household's own bind sends none), so `--accept-small-disk` cannot reach an install from the public image today (`audits/build-release-4b-2026-11-15/E/small-disk-flag-on-the-public-image.md`). Earlier: OPEN — filed 2026-11-15 | the next installer ISO for (1) | (1) a final exit code for a refusal that no retry can cure; the first-boot unit stops retrying on it, shows the sentence once, and tells the hub the reason; (2) the hub stores `small_disk` at the enrolment and shows it on the box's page | CC | | **R-1400** | Monitoring & notifications | P4 | **Every household mail but one prints the event's raw details as a line „Megjegyzés”.** Found 2026-11-15 while building the receipt mail of D84: the hub's household frame adds the event's details JSON under „Megjegyzés” (a note); the new receipt mail leaves it out, the other household mails keep it. A household reads a line of field names and values. Not measured: which mails, and how it reads in an inbox (`hub/internal/notify/testdata/mail_goldens/`). | OPEN — filed 2026-11-15 | — | read the household mail frame: which event types carry details worth a sentence, and drop the raw line for the rest | CC | | **R-979** | Box system & updates | P3 | **Controller updates are trusted by registry tag alone (no digest, no signature), and the Update button offers any newer tag, even an unvouched one; the built-in traefik, cloudflared and FileBrowser images are also pinned by tag only.** Whoever can push to the registry decides what runs. Fix: pin by digest from the hub's vouched list; offer only vouched versions. Security review 2026-10-12 (`audits/security-review-2026-10-12/`); detail held off-git for the operator (`/mnt/5_hdd/felhom.eu/evidence/security-review-2026-10-12-PRIVATE/`). | OPEN — design-proposal **2026-11-19 (security review 2, S2-29): still open as written in controller `10df99b8` — tag only, highest tag offered, the „pushed once” rule lives only in the build script.** | — | Design note, then build | CC | | **R-974** | Box system & updates | P3 | **A compromised host agent can likely become root on the host through the root wrapper that applies PBS settings: the wrapper accepts a folder path inside the agent's own writable tree and then acts on it as root.** Static (the symlink behaviour was reasoned, not run). Same class as R-861. Fix: a fixed root-owned folder, and refuse symlinks before any root action. Security review 2026-10-12 (`audits/security-review-2026-10-12/`); detail held off-git for the operator (`/mnt/5_hdd/felhom.eu/evidence/security-review-2026-10-12-PRIVATE/`). | OPEN **2026-11-19 (security review 2): not re-read in this pass; three more root helpers of the same class are in R-1428.** | — | Fix the wrapper with a test on a throwaway box | CC | @@ -348,7 +348,7 @@ stopping line that lies. | **R-1363** | Monitoring & notifications | P3 | **A household's mail ends with raw JSON: `Megjegyzés: {"avail_gb":5.192043304443359,"band":"warning",...}`.** Seen in the walks of 2026-11-02 (`audits/security-2026-11-02/part5/mail-fill-household.txt`, `mail-crash-household.txt`). `api/handler.go` adds the details as the „Megjegyzés" line; a golden pins it (`mail_golden_test.go`), so it was once wanted. The sentence itself is right and Hungarian. | OPEN — filed 2026-11-03 | — | drop the line from a household's mail, or render known fields in words; the operator's mail keeps it | CC | | **R-1365** | Monitoring & notifications | P3 | **A security-watch source that stays down mails once and never again; after that only one line of the Monday summary says so.** An open advisory reminds every 7 days, a failed source does not (`08` §11 [DESIGN], „once per streak"). The cost was seen 2026-10-28 → 11-02: one mail, then a week of silence. With the seven pins that can never be read now a stated GAP (R-1334), a streak means a real outage. | WAITING-ON-OPERATOR — D66 | D66 | on yes: `plan()` mails a streak again every 7 days, with a test seen failing; `08` §11's line changes | operator (decision), CC (build) | | **R-1371** | Monitoring & notifications | P3 | **Seven places have no space alarm, and the host-disk alarm that does exist goes to the household.** Found 2026-11-10 (the space list, `documentation/audits/space-list-2026-11-10/TABLE.md`, question 6). (1) **The guest root disk has no watcher.** The controller watches `/` and the system data path, but inside its container both are the data volume (measured on all three boxes, `evidence/extra/controller-fs-view.txt`); the 32 GiB guest root disk, which holds `/opt/docker/stacks`, is not a target. Closed row R-521 says it is — true only before the two volumes were merged; capability-map row C13 („warned before a filesystem fills — per filesystem”) does not hold for this one filesystem (the map's status is not changed here). (2) **The same filesystem is two targets**, and warnings are de-duplicated by path, so two warnings for one full disk are expected (read from source, not proved live; the comment at `cmd/controller/main.go:2694-2707` says the opposite). (3) **`host_disk_critical` and `storage_fill_critical` are mailed to the household by default** (`hub/internal/claim/engine.go:229-243`), but only the operator can free a host disk; the seed list was chosen as „critical only” and the question „can this reader act on it” was not asked. No architecture document names the host-disk event (search S4; `08-alarm-ladder.md` names the thin-pool one). (4) Small places with no alarm: EFI partition, temp folders, guest system log, guest package cache. **Members (answers marked N):** H10-6, H15-6, G1-6, G13-6, G15-6, G16-6, G17-6. | OPEN — filed 2026-11-10 (space list) **2026-11-10 12:22 — RULED (D68 = operator only; `09` §3 decision 287):** the household is not mailed about a host disk it cannot free. **2026-11-10 — parts (2) and (3) FIXED IN SOURCE, NOT RELEASED.** (2) felhom-controller `844f19f1`: one fill target per filesystem (`system.FilesystemKey`; the container's `/` is an overlay with its own device number, so the identity is taken from a Docker-kept file — `documentation/audits/space-list-2026-11-10/evidence/extra/fs-identity.txt`); `TestR1371_*`, red seen three times; not proven on a box. (3) hub source, no build: `host_disk_warning` / `host_disk_critical` are operator-only and the critical is not seeded; `storage_fill_*` reaches a household only for `usb`, `local-dir`, `nfs`, `cifs` (`notify.operatorOnlyFor`; `08-alarm-ladder.md`); `TestD68_*`, red seen three times. Parts (1) and (4) stay open: who watches the guest root disk is a design point in `16-storage-physical-identity.md`. **2026-11-10 (afternoon) — two design points added** (`16-storage-physical-identity.md` §14.4): a drive in the backup role reports as `usb` or `local-dir`, so the household gets the `storage_fill` mail for a drive it must not clean by hand; the hub's first list for a household still holds `storage_fill_critical` until the box sends its own list (not measured on the running hub). Also noted: the closed row R-521 and `felhom-controller/controller/cmd/controller/r521_f7_rootfs_test.go:29,36` call the watched path „the guest's root filesystem”; it is the container's `/`, an overlay on the data volume. The guest's real root disk is the one part (1) says nobody watches. **2026-11-10 (release run) — parts (2) and (3) RELEASED** (`documentation/audits/release-space-fixes-2026-11-10/`). (2) controller 0.352.0, global floor 15:31 UTC: on scratch guest 9202 one fill sent TWO warnings on the version before and ONE on the new one (so the double warning is now seen on a box, not only read from source); Tester 1, demo-hp and demo-felhom each check 2 filesystems where they checked 3. (3) hub 0.180.0 live since 15:12 UTC, Synced/Healthy; proof = the `TestD68_*` tests and the deployed commit. **Not proven on the running hub:** that a household which saved its own alarm list before no longer gets the host-disk type — the pages that show a household's list refuse a token (by design, `05` §17) and no test alarm was made (`16` §14.4). **Still open:** (1) the guest root disk has no watcher (design point); (4) the small places; a real fill on a household's box. | — | Controller: add the guest root disk as a fill-watch target, de-duplicate targets by filesystem, correct the comment; red-proof with two paths on one filesystem. Hub: after the pick, route the two host-disk events. Docs: name the host-disk alarm and its reader in `08-alarm-ladder.md`. | CC | -| **R-1379** | Monitoring & notifications | P3 | **An end date exists and nobody is warned before it (18 answers).** Found 2026-11-11 (the End-dates list, class T1, `documentation/audits/end-dates-list-2026-11-11/TABLE.md`; dates measured read-only on three boxes and from public certificates). No alarm fires before any of these ends: the box's web certificate (DB01-5, DB01-7: Traefik renews by itself; a renewal that fails is silent until the day), the hub address certificate (DH20-5), the operator tokens with a 90-day life (DH22-5), a node's Tailscale key (DH30-5), Debian's package keys (DB16-5), the 30-day restore copy (DB24-5), the end of OS security support (DB17-5), app and tunnel-client images (DB18-5, DB19-5), the local-API certificate (DB03-5), and the outside accounts: mail key, domains, Cloudflare, Hetzner, GitHub and Gitea tokens (DH23-5, DH24-5, DH27-5, DH28-5, DH29-5, DH31-5, DH32-5). Decision D77 in `STATUS.md` asks if the product warns 14 days before. **Ruled 2026-11-11 13:17: yes, D77 (`09` §3 decision 298). Not built.** | NARROWED — **2026-11-15 (build of release 4): D85 is BUILT, not released** (hub `main` after 0.182.0; `08` §12.3): once a day the hub reads the end date of each name in the setting `HUB_OWN_NAMES` and warns 14 days before, per name; a name that does not answer is its own line; with no setting the hub makes no call. **The release must add the setting to the manifest** (the lines are in `08` §12.3; the operator confirms the five names). No outside name was read. Earlier: NARROWED — **2026-11-15 12:49 — RULED (D85 = yes; `09` §3 decision 310).** Fix shape: once a day the hub opens a TLS connection to each of its five public names (`08` §12.2 row DH20; the list is configuration), reads the end date, and raises the operator-only alarm 14 days before, per name; a name that cannot be reached is a line in the daily ledger, not an end-date alarm. Earlier: NARROWED — **2026-11-15: the two warnings are LIVE** (hub 0.182.0, controller 0.353.0). The three test boxes' reports carry their web certificate's end date and issuer (read from a copy of the live database: 2026-12-31, 2027-01-02, 2027-01-19); the hub's end-date check set its baseline at the start. **Not seen:** no real certificate or token 14 days from its end, so no warning was seen. The other sixteen stay a design table; D85 is not ruled. Earlier: NARROWED — the two ruled warnings are built, not released. **2026-11-14 (night build):** the hub warns the operator 14 days before an operator token ends and 14 days before a box's web certificate ends (the box reads the date from its own store and sends it in its report) — `08-alarm-ladder.md` §12.1. The other sixteen are a design table there (§12.2): one is small and waits for a yes (the hub's own five public names), one is a design question (a household's Cloudflare token), the rest stay the operator's list. Not seen on a box or a deployed hub. OPEN — filed 2026-11-11 (end-dates list) | — | Fix shape (ruled, not built): the box reads its web certificate's date, the hub reads its tokens' dates, one operator alarm 14 days before. The outside accounts stay a list the operator reads (`documentation/audits/end-dates-list-2026-11-11/README.md`). | CC | +| **R-1379** | Monitoring & notifications | P3 | **An end date exists and nobody is warned before it (18 answers).** Found 2026-11-11 (the End-dates list, class T1, `documentation/audits/end-dates-list-2026-11-11/TABLE.md`; dates measured read-only on three boxes and from public certificates). No alarm fires before any of these ends: the box's web certificate (DB01-5, DB01-7: Traefik renews by itself; a renewal that fails is silent until the day), the hub address certificate (DH20-5), the operator tokens with a 90-day life (DH22-5), a node's Tailscale key (DH30-5), Debian's package keys (DB16-5), the 30-day restore copy (DB24-5), the end of OS security support (DB17-5), app and tunnel-client images (DB18-5, DB19-5), the local-API certificate (DB03-5), and the outside accounts: mail key, domains, Cloudflare, Hetzner, GitHub and Gitea tokens (DH23-5, DH24-5, DH27-5, DH28-5, DH29-5, DH31-5, DH32-5). Decision D77 in `STATUS.md` asks if the product warns 14 days before. **Ruled 2026-11-11 13:17: yes, D77 (`09` §3 decision 298). Not built.** | NARROWED — **2026-11-19 (release 4): the own-names part is LIVE** — hub 0.183.0 with `HUB_OWN_NAMES` (the five names, the operator's yes 2026-11-19): the start line names the five; the first daily read, 2026-11-20 06:40, read all five (the nearest end: `files.felhom.eu` 2026-12-23, 32 days — the warning is due 2026-12-09 if it is not renewed before). Earlier: NARROWED — **2026-11-15 (build of release 4): D85 is BUILT, not released** (hub `main` after 0.182.0; `08` §12.3): once a day the hub reads the end date of each name in the setting `HUB_OWN_NAMES` and warns 14 days before, per name; a name that does not answer is its own line; with no setting the hub makes no call. **The release must add the setting to the manifest** (the lines are in `08` §12.3; the operator confirms the five names). No outside name was read. Earlier: NARROWED — **2026-11-15 12:49 — RULED (D85 = yes; `09` §3 decision 310).** Fix shape: once a day the hub opens a TLS connection to each of its five public names (`08` §12.2 row DH20; the list is configuration), reads the end date, and raises the operator-only alarm 14 days before, per name; a name that cannot be reached is a line in the daily ledger, not an end-date alarm. Earlier: NARROWED — **2026-11-15: the two warnings are LIVE** (hub 0.182.0, controller 0.353.0). The three test boxes' reports carry their web certificate's end date and issuer (read from a copy of the live database: 2026-12-31, 2027-01-02, 2027-01-19); the hub's end-date check set its baseline at the start. **Not seen:** no real certificate or token 14 days from its end, so no warning was seen. The other sixteen stay a design table; D85 is not ruled. Earlier: NARROWED — the two ruled warnings are built, not released. **2026-11-14 (night build):** the hub warns the operator 14 days before an operator token ends and 14 days before a box's web certificate ends (the box reads the date from its own store and sends it in its report) — `08-alarm-ladder.md` §12.1. The other sixteen are a design table there (§12.2): one is small and waits for a yes (the hub's own five public names), one is a design question (a household's Cloudflare token), the rest stay the operator's list. Not seen on a box or a deployed hub. OPEN — filed 2026-11-11 (end-dates list) | — | Fix shape (ruled, not built): the box reads its web certificate's date, the hub reads its tokens' dates, one operator alarm 14 days before. The outside accounts stay a list the operator reads (`documentation/audits/end-dates-list-2026-11-11/README.md`). | CC | | **R-1383** | Monitoring & notifications | P3 | **A job on a timer does not run, gives up, or loses its state at a restart, and nobody is told (23 answers).** Found 2026-11-11 (the Clock list, class W1, `documentation/audits/clock-list-2026-11-11/TABLE.md`; read in source, times measured read-only on three boxes). The four worst: the guest-power watchdog gives up with one journal line (CA07-7, CA07-8); a signed job that is refused or runs out stays silent, so an agent update never arrives (CA09-4, CA09-7); the hub's daily checks are skipped with no trace when the hub is down at their minute (CH02-4, CH02-8, CH11-4, CH16-4); an operator mail waiting for a retry is lost at a hub restart, and a household mail is never retried (CH04-4, CH04-7, CH04-8, CH03-7). The others: CA13-7, CC07-7, CC12-4, CC19-7, CC20-4, CH07-4, CH09-4, CH14-4, CH26-7, CH27-4, CH27-8. Beside R-1361 (a lost event) and R-1186 (the silent off-site skip); not the same gaps. **Ruled 2026-11-11 13:17: yes (`09` §3 decision 299). Not built.** | NARROWED — **2026-11-15: LIVE** (hub 0.182.0, agent 0.175.1 vouched). Seen on the running hub: the daily-check ledger has its first records, the lost-mail scan ran at the start („none”), the „Security profiles” cell is filled for the three test boxes, the „Kernel clean-up” cell shows the N100's night. **Tests only, not seen on a box:** a refused signed job told, a guest-power give-up told, a lost operator mail told, a skipped daily check told. Earlier: NARROWED — **2026-11-14 (night build, second part):** the two things release 2a did not tell are built, not released: the System page has a cell for the kernel clean-up (removed, kept, space gained) and one for the security-profile refresh (the kernel, how many profiles changed, any failed service it restarted) — `11` §5.11.1, §5.12. On the way a display defect was fixed: a clean-up hid a kernel that fell back the night before in the „Kernel step” cell. NARROWED — the four worst cases are built, not released. **2026-11-14 (night build):** the operator is told when a hub daily check's slot passed while the hub was down, when an operator mail was never sent, when a box did not take a signed job in two hours, when a box refused a signed job (with the cause), and when the guest-power watchdog gave up (`05` §16.6, `08-alarm-ladder.md`). Tests only; not deployed, not seen on a box. LEFT: the other 15 answers of the class; a household mail is still never retried (R-1361); the kernel clean-up and profile-refresh detail (next part of release 3). OPEN — **2026-11-14 (release 2a), two more things the hub does not tell:** (1) a kernel clean-up that removed kernels shows only as „cleaned