skill-test

To validate YAML frontmatter and phase structure, you check required frontmatter fields, phase structure, and verdict keywords across skill directories. This structural compliance linting detects format and protocol gaps to maintain consistent studio workflows.

Behavioral testing for skill instructions evaluates skill logic against assertions defined in spec files. This spec-driven approach verifies that instructions meet expected behaviors and pass defined test assertions within the testing framework.

To run a coverage audit, you generate audit reports listing which skills, specs, and agent specs exist. The audit identifies exactly what has or has not been tested yet, ensuring full test coverage across the directory.

Skill validation works for CI-like maintenance workflows by checking one skill or the full set at once. It applies deterministic rule-based validations to ensure studio behaviors remain consistent before agents rely on them.

The best way to check skill quality is evaluating skills against category-specific quality metrics defined in the quality rubric. This category rubric evaluation ensures skills meet required standards for their designated category.

Skill testing fails when required frontmatter fields are missing because static linting detects these format and protocol gaps. Missing frontmatter breaks structural compliance, preventing skills from passing deterministic rule-based validations.