test-driven-development

Write failing tests before implementation by creating an executable specification for the desired behavior. You write a failing unit or integration test, implement the minimum code to make it pass, then refactor without breaking the suite.

Use the Prove-It Pattern for regression testing by writing a minimal failing test that reproduces the bug. Verify it fails on main, fix the root cause, then confirm the test and full suite pass.

The red-green-refactor cycle in unit testing starts with a failing test, builds the smallest change to make it pass, then improves code structure while keeping the test suite green.

Yes, behavior-based assertions require strong test naming conventions. This discipline favors refactoring-resistant assertions focused on behavior over implementation details across unit, integration, and end-to-end coverage.

Completion criteria for a test-driven development workflow require the full test suite to pass, such as pnpm test succeeding, and regression coverage must be added for any bug fixes.

You should limit end-to-end tests in the test pyramid by applying them selectively. Favor behavior-focused unit and integration tests to maintain a refactoring-resistant suite without excessive end-to-end coverage.