How do you run Playwright reliably in CI?

A reliable CI job has deterministic setup, a browser install step, a test command and artifact collection. The most common failures are missing system libraries, a forgotten browser install, an app that is not ready, and reports that disappear when the job fails.

name: Playwright
on: [push, pull_request]
jobs:
  test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 22
      - run: npm ci
      - run: npx playwright install --with-deps
      - run: npx playwright test
      - uses: actions/upload-artifact@v4
        if: always()
        with:
          name: playwright-report
          path: playwright-report/
          retention-days: 7

CI checklist

• Install browsers with --with-deps so OS libraries are present
• Use CI=1 and disable test.only with the forbidOnly setting
• Start the app before tests, through webServer or an explicit readiness check
• Upload the HTML report, traces and screenshots on failure
• Keep retries low and track flaky tests rather than ignoring them
• Use a fixed worker count that matches the runner's CPU and memory

If the pipeline runs on a small runner, a high worker count can cause timeouts that look like application bugs. Measure with a realistic worker count and memory limit. Also pin the browser version through the installed Playwright package so local and CI runs use the same engine.