CI Integration

GitHub Actions

Add one step to your existing workflow. The runner downloads automatically and figures out what to test from your repo — no per-language configuration needed.

name: CI

on: [push, pull_request]

jobs:
  test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Run tests via TestQuorum
        uses: testquorum/action@v0.1.0
        env:
          TESTQUORUM_API_KEY: ${{ secrets.TESTQUORUM_API_KEY }}

What the runner detects

The runner checks your repository for known manifests and turns on the matching test manager automatically:

• Cargo — detected from Cargo.toml. Enumerates every #[test] and doctest via cargo metadata and runs each individually.
• Nix — detected from flake.nix. Evaluates your checks attrset and builds each derivation as a separate check.
• treefmt — detected from treefmt.toml. Runs treefmt --ci.

Multiple managers can be active in the same repo at once — each runs independently. To tweak behaviour (point Nix at a different attrset, disable a manager, use a non-root Cargo.toml) add a .testquorum.tomlto your repository root. Most repos won't need one.

Pinning the action

The latest release is v0.1.0. To pin to a specific commit and have Renovate keep it current automatically:

- uses: testquorum/action@<commit-sha>  # kept current by Renovate

Submitting results directly via the API

If you are not using GitHub Actions, results can be submitted directly to the TestQuorum API. The endpoint is POST https://api.testquorum.dev/repos/{repo_id}/test-results where repo_id is your repository identifier (e.g. github:12345).

Each request body is a JSON object with a results array. Each entry describes a single test instance identified by a UUIDv7 you mint once and reuse across its lifecycle transitions:

POST /repos/github:12345/test-results
Authorization: Bearer <api-key>
Content-Type: application/json

{
  "results": [
    {
      "id": "0191e2a0-...",
      "test_name": "my_crate::tests::it_works",
      "test_manager": "cargo",
      "run": {
        "head": { "sha": "abc123", "height": 42 },
        "kind": { "kind": "post_land", "branch": "main", "trigger": { "kind": "push" } }
      },
      "state": {
        "kind": "passed",
        "discovered_at": 1749379200,
        "started_at": 1749379201,
        "finished_at": 1749379203,
        "duration_ms": 1847
      }
    }
  ]
}

Up to 1,024 results can be submitted per request. The full schema — including all state variants (discovered, running, passed, failed, errored, skipped, cancelled) and run kinds (diff, land, post_land) — is in the API reference.

Tests Dashboard

We're adding a Tests Dashboard feature — a GitHub issue on your repository rebuilt automatically that shows which tests are broken on main and a directory of every tracked test grouped by test manager (cargo, nix, treefmt). This is coming soon.

Next steps