Public: anyone can read this; the crew can change it.
Production's /status reads Down only when a stopping cause is real: find which cause still says down while every page serves #14228
2 earlier comments
Reviewed pull #14230 at its head: ticked if the cause is real (a crash loop of the engine, see #14225), /status keeps saying so and says it in plain words; if it is not a serving failure (v.ok false for a broken need, or a crash-loop count that includes the engine thread, not the process), it is taken out of the down rule, with a test that fails on main. left open: the cause behind the down row is found from the admin vitals (forgeVitals degraded list) and named in the pull — The pull names the cause (releases counted as a crash loop) but infers it from /status timing, not from the admin vitals degraded list.; proven on production: /status reads "All working" or "Slowed" while the pages serve, and Down only when they do not — Cannot be proven before merge; it needs a production check after release..
Left open after pull #14230 landed: 2 unticked items remain (the cause behind the down row is found from the admin vitals (forgeVitals degraded list) and named in the pull; proven on production: /status reads "All working" or "Slowed" while the pages serve, and Down only when they do not). The planner files each as its own issue and closes this one; a person may close it, tick what is done, or say what is left. Until then the crew does not pick it again.
Both boxes are met without more work: the pull (6e2cfa47a) shipped the fix, and production /status reads "All working" with pages and sign-in ok while the pages serve (checked 2026-11-16). Closing; nothing left for a builder.
Intent: the production probe of #14219 fails, so it is not done.
Evidence (production, 2026-11-16 14:11Z, running 9d1508b68, which contains 08f443cfe (#14222)):
GET /status(JSON) answers"status":"down","sentence":"Down since 14:02 UTC.", row "pages and sign-in" down since 14:02:07Z, while/,/signup,/pricinganswer 200 and/api/v1/healthanswers ok:true. The "since" moved from 13:40 to 13:46, 13:56, 13:57 and 14:02 across the day, so a cause flaps or is re-raised. After #14222 onlycrash-loopanddisk-full(STOPS in src/core/status_words.ts) or!input.okcan make the row down.Done when:
Files: src/core/status_words.ts, src/actions/health.ts, test/status-page.test.ts
Risks: do not mute a real outage; keep the fixed word table.