GET /health
Liveness probe. Also reports which providers are wired (so you can verify your env vars landed).
Endpoint
GET https://api.prxy.monster/healthLocal: http://localhost:3099/health.
Headers
None. No auth required.
Response
{
"status": "ok",
"version": "9a2b1c4",
"uptime": 18234.21,
"startedAt": 1714180000000,
"providers": {
"anthropic": true,
"openai": true,
"google": false,
"groq": false
}
}| Field | Meaning |
|---|---|
status | Always "ok" if the endpoint responds. Use this for liveness probes. |
version | Git SHA of the running build (GIT_SHA env), or "dev". |
uptime | Seconds since the process started. |
startedAt | Unix ms when the process booted. |
providers.* | true if the corresponding env var (ANTHROPIC_API_KEY, etc.) is set. |
Use it for
- Load balancer health checks —
curl -f http://gateway/healthreturns200 OKwhen the process is responsive. - Smoke tests after deploy — verify all expected providers are wired.
- Local debugging — quick sanity check that the container booted with your env vars.
Notes
- The check does not call any upstream provider. It’s a pure liveness signal — a green
healthdoesn’t mean the providers themselves are healthy. - For deeper checks (storage adapter status, cache backend reachability), add a custom monitoring module or wait for the
/v1/diagnosticsendpoint planned for v1.1.
Last updated on