Skip to content
Marriska Docs

Running tests in GitHub Actions

This guide gets marriska-runner running on every push and pull request from a GitHub Actions workflow. About 5 minutes of clicking through GitHub UI, plus a YAML file you commit to your repo.

Steps

Section titled “Steps”

  1. Create an API key

    Open Settings → Security → API Keys in Marriska and click Create API Key. Copy the value out of the dialog before you close it — it’s shown once, never again.

  2. Store the key as a GitHub secret

    In your GitHub repo, go to Settings → Secrets and variables → Actions → New repository secret. Name it MARRISKA_TOKEN and paste the key.

  3. Add the workflow

    Create .github/workflows/marriska.yml:

    name: Marriska tests
    on:
      pull_request:
      push:
        branches: [main]
    

    jobs:
      test:
        runs-on: ubuntu-latest
        steps:
          - uses: actions/checkout@v4
          - uses: actions/setup-node@v4
            with:
              node-version: "20"
          - run: npm install -g @marriska/runner
          - run: npx playwright install chromium --with-deps
          - run: marriska-runner run --test-set "Smoke Tests" --ci
            env:
              MARRISKA_TOKEN: ${{ secrets.MARRISKA_TOKEN }}

    The --ci flag tells the runner to skip auto-opening the report locally — exit code 0/1 drives the build status.

  4. Commit and push

    The workflow fires on the next push or PR. Watch progress in the Actions tab; the same run appears in your Marriska Reports page with full screenshots and timing.

Picking what to run

Section titled “Picking what to run”

The runner accepts either a test set or a single test. Use the test-set name or its UUID — both work.

- run: marriska-runner run --test-set "Smoke Tests" --ci
  env:
    MARRISKA_TOKEN: ${{ secrets.MARRISKA_TOKEN }}

Multi-browser runs

Section titled “Multi-browser runs”

Comma-separate the browsers; install the matching Playwright bundles in the same step that installs Chromium:

- run: npx playwright install chromium firefox webkit --with-deps
- run: |
    marriska-runner run \
      --test-set "Regression" \
      --browsers chromium,firefox,webkit \
      --parallel \
      --ci
  env:
    MARRISKA_TOKEN: ${{ secrets.MARRISKA_TOKEN }}

--parallel runs the browsers concurrently per test instead of sequentially. Useful when you’ve got a few minutes to save and your runner has the cores.

Saving the HTML report as an artifact

Section titled “Saving the HTML report as an artifact”

By default the runner writes an HTML report to ./marriska-reports. You can publish it back to the workflow run for offline review:

- name: Run tests
  run: marriska-runner run --test-set "Smoke" --ci
  env:
    MARRISKA_TOKEN: ${{ secrets.MARRISKA_TOKEN }}


- name: Upload report
  if: always()    # publish even on failure
  uses: actions/upload-artifact@v4
  with:
    name: marriska-report
    path: ./marriska-reports

if: always() is the important bit — without it, a failed run skips the upload and you lose the artifact.

Notifying on failures

Section titled “Notifying on failures”

Notification recipients live on schedules, not on the CLI runner. For CI failures, GitHub’s own status checks block the merge; if you want a separate alert channel, route the failed-workflow webhook through GitHub Actions notifications or GitHub’s email-on-failure preferences. Email-based test-set notifications from inside Marriska are configured per schedule — see Scheduling tests.

Section titled “Related”