Evidence stays bounded
The browser rejects missing, malformed, stale, or expired artifacts. A successful old observation never remains operational forever.
>_ public status / sanitized evidence
Status,
with receipts.
A deliberately bounded public surface. It consumes fresh, sanitized current-status observations from a GitHub-hosted runner outside this native cluster; separate rolling-window SLO evidence is published for reliability review.
Current state: Waiting for the latest external observation. This page accepts only fresh, sanitized evidence and never infers health from its own response.
01Component
Component
signals.
These are the user-facing services covered by the hourly external observation: portfolio, Pong, and analytics.
- Component evidence will appear with the next fresh observation.
02Evidence
Evidence
links.
Every published status.json observation carries bounded links to its workflow and history; rolling SLO evidence remains a separate artifact.
- Evidence links will appear with the next fresh observation.
03Incident
Incident
record.
Failures require confirmation; recovery requires consecutive passing observations.
- No incident record has been supplied.
04A small
A small
contract.
The checked-in fallback is safe by default; live claims require fresh, policy-bound evidence, and measured uptime includes its observation count.
Evidence travels with claims
GitHub Actions runs the checks outside this native cluster and commits sanitized history. The monitoring policy is human-approved; individual observations are automated. The publisher also generates slo.json for the internal 99%/30d objective, but incomplete coverage remains non-reportable and never becomes a public uptime claim.
Recovery is documented elsewhere
Backup and restore evidence is deliberately kept on the reliability evidence ledger. It is not current service status, and the ledger labels verified uploads, configuration declarations, and outstanding retention or rehearsal evidence separately.
05 / BOUNDARYOne cluster.
One cluster.
Honest limits.
Read the hourly monitoring policy and preserved evidence history.
Read the policy