| name | playwright-testing |
| description | Write end-to-end browser tests using Playwright. |
New E2E tests should be created in apps/prairielearn/src/tests/e2e. Each test file should be named *.e2e.spec.{ts,tsx} and contain tests for a specific feature or user flow.
expect and test should be imported from apps/prairielearn/src/tests/e2e/fixtures.ts. These include fixtures to start the PrairieLearn server and provide an editable test course.
Prefer using selectors that use user-facing content, such as getByRole, getByText, and getByLabelText. If that's impractical, consider using getByTestId instead. Avoid using locator() with implementation details like CSS classes or IDs unless absolutely necessary.
Playwright e2e tests run fully in parallel (fullyParallel: true). Each test must be self-contained — set up its own state (users, enrollments, etc.) rather than relying on shared mutable state from beforeAll. Use the courseInstance fixture for the default course instance. Don't use module-level variables to share state between tests.
Conventions for writing tRPC routers and procedures in PrairieLearn.
Backwards compatibility rules when changing element controllers in `apps/prairielearn/elements/`.
Keeping the AI question generation HTML validator in sync with element implementations.
Dealing with groups and teams in PrairieLearn, and confusion about table/column/variable names.