After this, the AI writes a failing test that reproduces the desired behaviour BEFORE writing any production code — bugs caught at write-time, regression coverage as a by-product. For developers who agree TDD is correct but can't get the AI to do it without reminding it. Before: "Add validation: password must be 8+ chars with one number" [Writes the validation function directly. Looks fine. Three weeks later you discover passwords with only numbers and no letters pass. No test, no signal.] After: "Add validation: password must be 8+ chars with one number" Test first: expect(validate("pass")).toBe(false) // too short expect(validate("password")).toBe(false) // no number expect(validate("password1")).toBe(true) // valid expect(validate("12345678")).toBe(false) // numbers only — the gotcha Runs tests — three pass, one fails (the gotcha catches it). NOW writes the production code to make all four pass. Use when starting a new feature or bug fix "TDD this" an existing change has no test covering it Three ways to use this Paste it (any AI, 30 seconds, no install)Open the .md file, copy contents, paste into Claude / ChatGPT / Cursor / Perplexity with: "Read this and apply it whenever the task it describes comes up in our conversation." Done. Install it (Claude Code / Cursor users)Drop the unzipped folder into ~/.claude/skills/ (Claude Code), .cursor/skills/ (Cursor), or your IDE's equivalent. Auto-loads next session. New to AI assistants? Get Claude free at claude.ai, then use Option 1. €7
