What it does
A skill that makes AI coding agents follow a test-driven workflow. The agent writes a failing test, writes the code that makes it pass, then refactors, and it checks the requirements before it starts. Each coding task also gets a peer review step. It works with Claude Code, Codex, Cursor and Windsurf, and pairs with companion tools for code navigation and web search.
Use cases
- 01Add a feature with its tests written first
- 02Set the requirements before the agent edits code
- 03Review agent-written changes before accepting them