2026-11-18 stocktake: re-verify 7 stale pages against live state, resolve gmail-digest refire incident

This commit is contained in:
Tony0410
2026-11-18 03:19:54 +08:00
parent 4574c4ebc3
commit 98610466cc
11 changed files with 199 additions and 20 deletions
+12
View File
@@ -1844,3 +1844,15 @@ recorded here so the next audit does not re-litigate it.
(`status: active`, `review_after` 2026-11-18); the incident-signature fragmentation fix remains
**proposed, not applied**; the plugin-load/hook-budget health check is **proposed, not applied**.
## 2026-11-18 03:xx — 03:00 stocktake (cron)
- Ran `wiki-maintain.sh`: skills-index regenerated (299 skills / 67 categories, content unchanged vs 11-17), freshness report `raw/stocktake/freshness-20261118.txt` — 9 stale.
- Live-vs-page re-verification of stale pages; 7 re-verified with dated corrections:
- `runbooks/provider-health.md` — snippets re-executed: CT600 /models 200 (6,501), omni-101 /api/health 200, fallbacks 200.
- `runbooks/update-hermes-safely.md` — prerequisites re-verified (v0.23.0, venv + managed runtimes present).
- `incidents/2026-10-03-ct450-hermes-update.md` — CT450 now v0.23.1 (0 behind origin/main), CT460 v0.23.0; CT255 mattermost running + peer online → Mattermost-timeout claim superseded.
- `incidents/2026-10-03-mnemosyne-vector-recall-fts-only.md` — coverage monitor ok/streak 0; fix holds (script blocked by lifecycle guard; verified via monitor).
- `incidents/2026-10-25-gmail-digest-runaway-and-refire.md` — RESOLVED: last_status ok, one clean daily run 11-12→11-17, no refires; delivery_queued cleared.
- `incidents/2026-10-16-omniroute-503-resource-pressure.md` — 22 days clean, stays monitoring.
- `incidents/2026-10-20-gateway-unclean-exit-orphaned-run.md` — no unknown rows in retained window; 10-20 log line aged out (rotation noted on page).
- Left stale (flagged only, not verifiable this run): `reports/communities-wa-social-media-audit-2026.md` (24d), `incidents/2026-10-24-conduit-call-request-refused.md` (23d, open behavioural).
- Quartz rebuilt + deployed via second `wiki-maintain.sh` run (pattern 13).