Checkpoint 6.7: public status, inside the app

GET /status.json serves the sanitized artifact unmodified — outside /api, no token — and /status renders it; stale or missing reads 'monitoring unavailable', never fabricated rows.

at a glance

Goal: Public viewers see the sanitized snapshot without ever touching management APIs.
Built: GET /status.json mounted outside /api (no token) serving the same file the snapshot writer produces — hashed aliases, self-expiring — plus a /status page that renders it.
Result: three Playwright specs: live snapshot, stale doctor, missing file — the last two read 'monitoring unavailable', never 'all down'
Code: akira-shirase-oss/monitoring-service · main @ dcd2ac2

Seventh slice.

Gap: the sanitized status existed only as the separate static app; inside the dashboard there was nothing for a viewer without the token.

What changed: GET /status.json serves the same sanitized artifact unmodified — mounted outside /api, no token, hashed service aliases — and /status renders it in the app’s own components. The expiry contract is preserved end-to-end: stale or missing reads “monitoring unavailable” with the last-seen evidence, never fabricated rows and never “all down”. The artifact IS the contract — the page doesn’t re-derive status from management data.

Proof: three Playwright specs — a live snapshot renders its rows, the stale-doctor fixture flips the verdict to “monitoring unavailable” while rows stay as stale evidence, and a missing file yields the same verdict with no rows at all.