playwright-testing

name	Playwright Testing
description	Write end-to-end Playwright tests that are resilient, readable, and free of flakiness.

Playwright Testing

Tests should fail only when the app is broken — never because of timing.

Selectors

• Prefer user-facing locators: getByRole, getByLabel, getByText.
• Use data-testid only when semantics aren't enough.
• Avoid brittle CSS/XPath tied to layout.

No flaky waits

• Never waitForTimeout with a magic number.
• Rely on Playwright's auto-waiting and web-first assertions (await expect(locator).toBeVisible()), which retry until true.
• Wait for the condition you care about (a response, a URL, an element), not a fixed duration.

Structure

• One behavior per test; arrange-act-assert.
• Isolate state: reset/seed data so tests don't depend on order.
• Use fixtures for shared setup (auth, base URL).

Rules

• Make tests deterministic — mock network where the backend is out of scope.
• Name tests by user intent ("user can reset password"), not implementation.
• Run headed locally to debug, headless in CI with traces on failure.