← ClaudeAtlas

test-and-fixlisted

Run tests, analyze failures, and automatically fix them
iamnolanhu/claude-dotfiles · ★ 1 · Testing & QA · score 63
Install: claude install-skill iamnolanhu/claude-dotfiles
# Test and Fix Run the project's test suite, analyze any failures, and automatically fix them. ## Process 1. **Detect test framework** by checking for: - `bun test` (check for `bun.lockb` or test files with Bun imports) - `vitest` (check for `vitest.config.*`) - `jest` (check for `jest.config.*` or jest in package.json) - `pytest` (check for `pytest.ini`, `pyproject.toml` with pytest) - `go test` (check for `*_test.go` files) 2. **Run tests** using the detected framework: ```bash bun test # Bun projects bunx vitest run # Vitest bunx jest # Jest uv run pytest # Python go test ./... # Go ``` 3. **Analyze failures** by: - Reading the error messages and stack traces - Identifying which files and functions failed - Understanding the root cause (assertion failure, runtime error, etc.) 4. **Fix issues** by: - Reading the failing test file - Reading the implementation file being tested - Making the minimal change to fix the failure - Re-running tests to verify the fix 5. **Iterate** until all tests pass or you identify a deeper issue requiring human input. ## Guidelines - Fix the **implementation**, not the test (unless the test is clearly wrong) - Make **minimal changes** - don't refactor while fixing - If a test seems fundamentally wrong, ask before modifying it - If you can't fix after 3 attempts, explain the issue and ask for help