playwright-best-practices

Flaky Playwright tests usually stem from race conditions, shared state between tests, or missing auto-waiting assertions. Replace explicit waits with retrying assertions like expect(locator).toBeVisible(), ensure test isolation with fixtures, and use the trace viewer to inspect failures.

Page Object Model encapsulates page locators and actions into reusable classes, keeping tests readable and maintainable. The skill covers when to choose POM versus fixtures, how to structure test suites, and how to embed test steps inside page objects for better reporting.

Use API tests with the request fixture to validate status codes, response bodies, and business logic without launching a browser, which runs 10-100x faster. Reserve E2E tests for user-facing flows, error rendering, and navigation behavior.

Yes, Playwright integrates with @axe-core/playwright to run automated accessibility scans. You can analyze full pages or scoped regions, filter by WCAG tags, disable specific rules, and build a reusable fixture that fails tests on violations.

The skill includes dedicated references for GitHub Actions, GitLab CI, Docker, and other providers. They cover browser installation, sharding for parallel execution, artifact uploads for traces and reports, and retry configuration for CI environments.

Parallel-only failures typically indicate shared mutable state, port conflicts, or test data collisions between workers. Use worker-scoped fixtures, unique test data per worker, and proper isolation instead of depending on execution order.