The healthcheck ran 'curl -f', but the Emby image has neither curl nor a
standalone wget -- only BusyBox. The probe ENOENT'd every run, the container
stayed permanently unhealthy, and Traefik will not route to an unhealthy
container, so the customer got a 404 while Emby itself was serving fine.
Switched to the BusyBox-wget family via the /bin/busybox multi-call binary.
Campaign 7 catalog sweep.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nn3VgQk9iwEGgyx6QJ2NvE
The frontend healthcheck ran 'wget --spider', but the image is distroless: no
shell, no wget, no curl, and node itself is only reachable at an absolute path
(/nodejs/bin/node, off PATH). The probe therefore ENOENT'd on every run, the
container stayed permanently unhealthy, and Traefik -- which will not route to
an unhealthy container -- returned 404 to the customer even though the app was
serving 200 on :3000 the whole time.
Replaced with the Node-exec family using the absolute interpreter path.
Live-verified on the demo box.
Campaign 7 catalog sweep.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nn3VgQk9iwEGgyx6QJ2NvE
MAJOR: immich v3.0.0 drops pgvecto.rs support and requires VectorChord. Our pin
was already VectorChord, so fresh deploys are unaffected; an in-place update from
a pgvecto.rs-era install would need the upstream migration first.
Postgres sidecar moved to the vectorchord/pgvectors extension versions immich
v3.0.3 ships in its own compose (0.4.3 / 0.2.0), keeping our PG major (16)
rather than upstream's 14 to avoid a needless major change.
NOTE (recorded, not fixed): upstream v3 migrated redis -> valkey:9. Kept
redis:7-alpine here; the swap is a structural change, not a pin bump.
Campaign 7 catalog sweep.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nn3VgQk9iwEGgyx6QJ2NvE
MAJOR (bookstack): v26.05 needs storage/fonts writable for PDF export; revision
viewing is now a separate permission. Fresh deploys unaffected; flagged for the
future in-place update path.
MAJOR (mariadb): 11.6 -> 12.3.
Campaign 7 catalog sweep.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nn3VgQk9iwEGgyx6QJ2NvE
Adds the referential-coupling backup: block to every app that binds
${HDD_PATH}/${USERDATA_PATH} (immich, paperless-ngx, nextcloud, calibre-web,
audiobookshelf, komga, navidrome, radarr, sonarr, emby, jellyfin, plex, romm).
Each lists its userdata:/hdd: binds with class mandatory|optional|excluded
(operator-ruled + spike SQ2).
Requires controller v0.132.0 (deployed) which parses+validates these blocks.
INERT — no backup tier changes behavior yet (Task 3/4 consume it). All 13
verified against the shipped parser: parse-clean, every bind resolves explicit.
audiobookshelf media/audiobooks = optional (PENDING Viktor veto to excluded).
Ref: felhom.eu/documentation/audits/SPIKE-backup-classification-2026-07-14.md
bentopdf :latest -> v2.8.6; calibre-web :latest -> v4.0.6 (== running digest on
demo 9201, c31a738b - pin is a no-op); papra :latest -> 26.6.1-rootless (latest
was the rootless variant); recipe-importer :latest -> v0.9.11 (tag pre-existed,
digest-equal, no retag needed); termix :latest -> 2.5.0.
All five pins digest-identical to what :latest resolved to on 2026-07-12.
New gate scripts/check-image-pins.py (catches floating tags AND untagged refs;
red-proofed both shapes). Standing rule in CLAUDE.md + REUSE.md row.
BusyBox wget (+ node/python/curl one-shots, incl mealie's socket tuple) resolve
localhost -> IPv6 ::1 with no cross-family fallback; an IPv4-only-binding app
reads docker-unhealthy while serving (vaultwarden, re-run 2026-07-06). Escalates
that instance to the class. Scoped strictly to healthcheck test: lines
(diff-reviewed: no env/config/label changed; .felhom.yml already clean). New
REUSE.md convention row.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PSK5g6qYLknKj8u3QAFEr6
Vaultwarden treats defined-but-empty SMTP_HOST/SMTP_FROM as "set"; with upstream
_enable_smtp defaulting true its validation errors out and the container
crash-loops on every fresh deploy with app-email off (the default). Gate the
SMTP group with _ENABLE_SMTP: compose default false (clean boot, mail off),
flipped "true" by the app-email injection via smtp_mapping.extra. Proven on the
pinned 1.33.2-alpine image (P1 repro exit 12 / P2 off boots / P3 on boots).
REUSE.md: strict-image TRAP row. Sweep note: the other 5 smtp-mapped templates
are campaign-boot-proven tolerant; no edits.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PSK5g6qYLknKj8u3QAFEr6
calcom + nextcloud both send live via the :2526 plaintext listener (calcom signup =
Nodemailer; nextcloud password-reset = Symfony Mailer), both From <app>@felhom.eu.
calcom healthcheck: wget GET not --spider (HEAD 405s). CHANGELOG + REPORT.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
/api/health doesn't exist in cal.com v4.x (404) -> container stayed unhealthy ->
Traefik wouldn't route it. /api/auth/providers is a stable 200 once serving.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Cal.com's prisma schema declares directUrl=env(DATABASE_DIRECT_URL); without it the
startup 'prisma migrate deploy' fails (P1012) and the schema stays incomplete -> the app
500s. Set to the same direct Postgres URL. Pre-existing, never caught (bad image pin
meant calcom never deployed).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
calcom/cal.com:v4.8.7 is a git tag with no published Docker image; the manifest is
unknown. v4.6.9 is the nearest valid 4.x image (same EMAIL_SERVER_* env behavior).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The wget healthcheck always failed (exit 127, no wget in image) → container
unhealthy → Traefik refused to route it. Pre-existing, never caught because the
3.12.1 image pin didn't exist.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
New .felhom.yml field consumed by the controller's Megnyitás link. Audit of all 53 templates: only
gokapi (file-share index redirects away) and ghost (bare / is the public blog; admin at /ghost/) need it;
the rest serve/redirect usefully at root. onlyoffice is API-only (status page at /) — left at root.
Gokapi's bare root '/' redirects to RedirectUrl. It was the upstream default (github.com/Forceu/Gokapi),
so the controller's 'Megnyitás' link (always the bare subdomain root) landed on Gokapi's GitHub instead
of the app. Point it at https://${SUBDOMAIN}.${DOMAIN}/admin so the root + the controller link reach
the Gokapi login/panel.
App now auto-configures (no setup wizard), so first_steps described a flow that no longer happens.
Add default_creds 'Felhasználó: admin · jelszó a Beállítások oldalon' so the app-info page shows the
initial admin user like other apps; password stays on the Beállítások page (generated field).
Gokapi v1.9.6 has no env-var setup and otherwise exposes a public /setup (admin-takeover risk).
Entrypoint seeds config.json on first boot (admin, this app's URL, local storage, no-encryption so
it restarts without a prompt) with password/salts cleared, then runs the documented
--deployment-password one-shot to set the felhom-generated admin password before serving — claims
admin at first boot, never exposes /setup. Seed pinned to v1.9.6 (ConfigVersion 21).
Verified against Gokapi docs + the v1.9.6 binary: no env-var headless admin setup exists in any
version. Non-interactive setup requires a pre-seeded config.json + --deployment-password. Reverting
the non-functional env so the template isn't misleading; choosing a real fix path with the user.
Gokapi sat in maintenance mode since deploy because /setup was never completed and the
template provided no headless-setup env. GOKAPI_USERNAME/PASSWORD configure it on first
start; password is felhom-generated (deploy field, shown to customer). Fixes share.* maintenance page.
Surfaced during the unattended deploy sweep: a fresh gitea reported unhealthy because
the HC hit /api/v1/version (404 until install-lock); /api/healthz returns 200 while
serving. Fixed both the docker HC and the controller .felhom.yml probe; bumped
start_period to 90s for first-boot DB migration. Same class as the komga fix.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
admin