Platform status
What this deployment is doing, without asking you to sign in.
The application is serving and the database is answering. One scheduled job is late: Scheduled job watchdog.
Degraded
- ApplicationServing. This page was rendered by it. /api/health is the same answer for a machine.
- DatabaseAnswering. Read at 15 Sept 2026, 05:06 UTC.
- ReleaseThis deployment did not record the commit it was built from.
- Recovery pointnpm run ops:recovery-evidence has not been run against the hosted project from this tree. Until it has, this deployment states a recovery point objective and evidences none.
The scheduled work
Each of these runs on a schedule this deployment does not control from inside itself. A job is late when its last completion is older than two of its own intervals, which is the same rule the watchdog applies every fifteen minutes.
Connected system sync
No pass recorded- Runs
- every 15 minutes
- Last completed
- not yet
- Next expected
- not yet
Work brought in from a connected system stops arriving.
Environment probe
On time- Runs
- every hour
- Last completed
- 15 Sept 2026, 05:00 UTC
- Next expected
- 15 Sept 2026, 06:00 UTC
If it stops: deployment environments stop reporting whether they are reachable.
Front door health check
On time- Runs
- every 5 minutes
- Last completed
- 15 Sept 2026, 05:05 UTC
- Next expected
- 15 Sept 2026, 05:10 UTC
If it stops: nobody is told when the sign-in page stops answering.
Pod stream mail and channel cards
On time- Runs
- every 15 minutes
- Last completed
- 15 Sept 2026, 05:00 UTC
- Next expected
- 15 Sept 2026, 05:15 UTC
If it stops: pod announcements and mentions stop leaving the building.
Presentation asset reconciliation
On time- Runs
- every 15 minutes
- Last completed
- 15 Sept 2026, 05:00 UTC
- Next expected
- 15 Sept 2026, 05:15 UTC
If it stops: deleted presentation files are left behind in storage.
Scheduled job watchdog
Late- Runs
- every 15 minutes
- Last completed
- 15 Sept 2026, 00:45 UTC
- Next expected
- 15 Sept 2026, 01:00 UTC
Nothing is watching the other jobs, which is the failure this list exists for.
Weekly resourcing snapshot
No pass recorded- Runs
- every week
- Last completed
- not yet
- Next expected
- not yet
The week opens with no resourcing figures and last week is never closed.
Keep the server function warm runs on a schedule and is deliberately not watched: its absence costs a slower first request rather than a missed obligation, and paging somebody for that teaches them to ignore the page.
The dates on the evidence
Every figure this product publishes about its own security carries the date it was taken. A security figure with no date is a claim about a moment nobody can identify, so the dates are here rather than only on the pages that quote them.
- Hosted measurement10 September 2026, 4 days ago. What the deployment enforces.
- Isolation proof10 September 2026, 4 days ago. A real session tried to read another tenant and was refused: pass.
- This release's schemaMigration ledger head 20260914010000, from 293 migrations. Regenerated on every push.
- Recovery settingsnpm run ops:recovery-evidence has not been run against the hosted project from this tree. Until it has, this deployment states a recovery point objective and evidences none.
This page is not a service level commitment, and this product does not hold one. It reports what was observed. The security page states what is implemented and what is not, in the same words. /status.json is this page for a machine, and it is the same reading rather than a second one.