CrabGlamp

Status

Real-time CrabGlamp service status. The full status page is hosted externally so it stays available during platform-side outages. This page summarizes which components have status pages, what each tracks, and where to subscribe to incident notifications. Update incidents in progress are also listed on the dashboard.

Last updated:

The CrabGlamp status page lives at status.crabglamp.com. It is hosted externally so it stays available when our control plane has problems.

What the status page tracks

The status page reports on each of these components independently:

  • Control plane — the dashboard, API, and webhooks at crabglamp.com.
  • Agent compute — Hetzner and Fly capacity for new Agent provisioning and existing Agent lifecycle.
  • Plain VM compute — Hetzner capacity for new Plain VM provisioning and existing VM lifecycle.
  • LLM proxy — the central proxy that handles virtual key authentication and provider routing.
  • Voice proxy — the same proxy's TTS/STT path.
  • App integrations broker — the OAuth broker for Google, GitHub, and Spotify.
  • Public glamp delivery — S3 (Tigris) availability for downloading published glamps.

Subscribing to incidents

The status page exposes:

  • RSS at status.crabglamp.com/feed.
  • Email subscriptions — sign up at the top of the page.
  • Webhook subscriptions — for incidents and resolved incidents. Configure on the page.

In-dashboard banner

When an incident affects the control plane or the LLM proxy, the dashboard at /dashboard shows a banner pulled from the same status feed. The banner clears automatically once the incident resolves.

Maintenance windows

Scheduled maintenance is announced on the status page at least 48 hours in advance. Customer-visible impact (failed provisioning, transient API errors) is called out explicitly.

Related

View as Markdown — the same content as plain text for AI assistants and offline reading.

Was this helpful?