writing-skills

Test-driven development for skill documentation uses a RED-GREEN-REFACTOR cycle to write, test, and verify skills before deployment. Start with baseline RED tests to capture failures, then iterate until the skill passes mandatory deployment verification.

The RED-GREEN-REFACTOR workflow is a structured cycle for skill creation. RED captures baseline failures and rationalizations, GREEN ensures the skill passes tests, and REFACTOR iteratively refines the documentation for robust deployment.

Skills often fail due to undocumented rationalizations and behavioral gaps. Capturing explicit rationalizations during baseline testing and applying meta-testing prevents these failures by ensuring the skill logic is fully verified before deployment.

Structuring a SKILL.md file requires specific frontmatter, clear sections, and progressive disclosure of information. The test-driven approach validates this structure through iterative refactoring and deployment checks to ensure quality assurance.

Yes, test-driven patterns support skill editing and refactoring across agent workflows. You apply the RED-GREEN-REFACTOR cycle to existing skills by establishing baseline tests, identifying red flags, and iteratively improving the documentation.

Test-driven skill writing requires mandatory baseline RED tests and explicit rationalization capture before proceeding. If you cannot define concrete failure modes or establish deployment verification checks, this structured approach cannot validate the skill effectively.