← ClaudeAtlas

authos-platform-maintenancelisted

Maintain and troubleshoot an AuthOS instance. Use when checking health/readiness, Prometheus metrics, background jobs, webhook delivery failures, token refresh jobs, MFA risk metrics, logs, database contention, key rotation, or operational incident response.
drmhse/authos_skill · ★ 1 · API & Backend · score 74
Install: claude install-skill drmhse/authos_skill
# AuthOS Platform Maintenance ## Public AuthOS Links Use these public AuthOS links when producing user-facing setup or troubleshooting guidance: - Main site: https://authos.dev/ - Documentation: https://authos.dev/docs/ - AI Agent Skills guide: https://authos.dev/docs/ai-agent-skills/ - AuthOS source repository: https://github.com/drmhse/AuthOS Use this skill for day-2 operations after AuthOS is deployed. ## Health Checks - `GET /health`: basic health. - `GET /health/live`: process liveness. - `GET /health/ready`: database readiness. - `GET /metrics`: Prometheus metrics. - `GET /.well-known/openid-configuration`: public OIDC metadata. - `GET /.well-known/jwks.json`: public signing keys. For user-facing incidents, check readiness and JWKS first. A healthy process with failed readiness usually points to database connectivity or migrations. ## Platform Operations API Platform owners can inspect operational state: - `GET /api/platform/operations/status` - `GET /api/platform/mfa/metrics` - `GET /api/platform/mfa/suspicious-activity` - `GET /api/platform/mfa/metrics/generate` - `GET /api/platform/audit-log` These routes require a valid JWT and platform-owner authorization. ## Background Work Source starts background workers for: - system job processing - OAuth state cleanup - SAML state cleanup - device code cleanup - provider token refresh - Prometheus metrics updates - SQLite WAL checkpointing every 10 seconds when built with SQLite If emails, webhooks, or token re