writing-skills

Test-driven skill documentation captures failing RED baseline tests before writing the minimal Skill, then refactors to close loopholes. This enforces a structured workflow with YAML frontmatter, rationalization tables, and a deployment checklist for reliable validation.

A RED baseline test captures agent rationalizations and expected failures before writing the Skill. Running it first ensures the minimal GREEN Skill and subsequent REFACTOR tightening directly address documented edge cases and deployment readiness criteria.

Validate deployment readiness by running structured workflow checks including YAML frontmatter verification, rationalization table review, red flag detection, and a final deployment checklist. This ensures the minimal Skill passes GREEN tests before release.

Yes, flowcharts support skills-testing by mapping agent decision paths and rationalizations. Combined with subagent tooling references, they help visualize and tighten Skill logic during the REFACTOR phase to close documentation loopholes.

Skill-writing operates with zero dependencies, using internal subagent tooling references for testing instead. The structured workflow relies on YAML frontmatter and rationalization tables rather than external libraries for validation.

A rationalization table documents agent reasoning during baseline tests, exposing edge cases and red flags. It provides structured traceability during the REFACTOR phase, ensuring the minimal Skill closes loopholes before deployment.