skf-test-skill

To validate skill documentation completeness, you can audit the SKILL.md structure and API surface coverage against code and references. This process checks cross-references and integration patterns to ensure the documentation supports reliable AI agent instructions.

Skill coherence testing verifies that a skill's documentation, API surface, and references align with the actual code. It compares SKILL.md to the actual surface and validates provenance data, returning a deterministic completeness score with evidence-backed results.

You generate a gap report with traceable file:line citations by running a completeness audit on your skill. The report identifies missing documentation or API mismatches, providing traceable citations for all findings to locate exact gaps.

You can test a single skill independently using naive mode, or validate an entire stack using contextual mode. Contextual mode supports cross-skill coherence validation, including external validators and a health-check workflow integration.

The best way to check if a skill is production-ready is to run an automated completeness audit that scores documentation, API surface, and cross-references. The audit returns a deterministic score with threshold handling and an atomic write of the result.

Your skill documentation scores low on completeness checks when its SKILL.md structure, API surface coverage, or cross-references fail to match the actual code and references. The generated gap report provides traceable file:line citations to identify specific missing elements.