If the current repo has its own rules/skills covering this topic (check .claude/rules/ and repo CLAUDE.md), those take precedence — apply this skill only where they're silent.
Writing Tests
Core principle: Test user-observable behavior with real dependencies. Tests should survive refactoring.
"The more your tests resemble the way your software is used, the more confidence they can give you." — Kent C. Dodds
Why this matters: Tests exist to give you confidence. The Testing Trophy prioritizes integration tests because they test real behavior across real modules — giving maximum confidence per test written. Unit tests in isolation often just test mocks, not your actual system.
Testing Trophy Model
| Priority | Type | When |
|---|---|---|
| 1st | Integration | Default - multiple units with real dependencies |
| 2nd | E2E | Complete user workflows |
| 3rd | Unit | Pure functions only (no dependencies) |
Mocking Guidelines
Default: Don't mock. Use real dependencies.
Only mock:
- External HTTP/API calls
- Time/randomness
- Third-party services (payments, email)
Never mock:
- Internal modules
- Database queries (use test DB)
- Business logic
- Your own code calling your own code
Before mocking, ask: "What side effects does this have? Does my test need those?" If unsure, run with real implementation first, then add minimal mocking only where needed.
Test Type Decision
Complete user workflow? → E2E test Pure function (no side effects)? → Unit test Everything else → Integration test
Assertion Strategy
| Context | Assert On | Avoid |
|---|---|---|
| UI | Visible text, roles | CSS classes, internal state |
| API | Response body, status | Internal DB state |
| Library | Return values | Private methods |
Anti-Patterns
| Pattern | Fix |
|---|---|
| Testing mock calls | Test actual outcome |
| Test-only methods in production | Move to test utilities |
sleep(500) | Poll for actual condition |
| Asserting on internal state | Assert on observable output |
| Incomplete mocks | Mirror real API completely |
Quality Checklist
- Happy path covered
- Error conditions handled
- Real dependencies used (minimal mocking)
- Tests survive refactoring
- Test names describe behavior
Language-Specific Patterns
- JavaScript/Typescript/React: See references/typescript-react.md
- Python: See references/python.md
- Go: See references/go.md
Async Waiting
Wait for the actual condition, not a guess about how long it takes.
typescript// Bad: arbitrary delay await new Promise(r => setTimeout(r, 2000)); expect(element).toBeVisible(); // Good: poll for condition await waitFor(() => expect(element).toBeVisible());
Prefer framework built-ins:
- Testing Library:
findByqueries,waitFor - Playwright: auto-waiting,
expect(locator).toBeVisible() - pytest:
asyncio.wait_for, tenacity
Language-specific waiting patterns:
- TypeScript: See references/waiting-typescript.md
- Python: See references/waiting-python.md
Remember: Behavior over implementation. Real over mocked. Outputs over internals.

