thump deploy gate — GitHub Action
Probe a thump monitor from every configured region inside your CI pipeline and fail the job when the target is down. Typical uses:
- Post-deploy verification — deploy, then gate the "promote" step on the
service actually answering from every region.
- Pre-merge canary — point a monitor at your staging/preview URL and gate
the merge on it being healthy.
Usage
- name: Verify service is up (thump)
uses: rbaljinder/thump/.github/actions/thump-check@main
with:
api-token: ${{ secrets.THUMP_API_TOKEN }}
monitor-id: 0f8b7c62-1234-5678-9abc-def012345678
# fail-on: majority # default; "any" is stricter
Inputs
| Input | Required | Default | Description |
|---|---|---|---|
api-token | yes | — | thump personal access token (Settings → API tokens). Store as a repo secret. |
monitor-id | yes | — | Monitor id — visible in the dashboard URL (/app/monitors/<id>). |
api-url | no | https://thump.dev | API origin; change for self-hosted instances. |
fail-on | no | majority | any = fail if any region is down; majority = fail when ≥ half the regions are down (mirrors thump's own quorum logic). |
What it does
Calls POST /api/v1/monitors/:id/run — the same "Run now" the dashboard button uses — which probes the target from every region the monitor is configured for, synchronously, without touching the monitor's schedule. The per-region results (status, latency, cause) are rendered into the job's step summary, and the raw JSON is exposed as the results output for downstream steps.
The probe run also feeds the monitor's regular history, so gated deploys show up in your check history like any other probe.