What problem does it solve? Reviewing pull requests that add or modify Agent Skills requires checking YAML frontmatter validity, naming conventions, description quality, and marketplace manifest consistency, which is tedious and error-prone when done manually. ## Core Features & Use Cases - Structural Validation: Mechanically verifies YAML frontmatter, required name and description fields, kebab-case naming, line limits, and path formats. - Design and Marketplace Checks: Evaluates description trigger quality, progressive disclosure, degrees of freedom, name collisions, and plugin.json consistency. - Verified Findings Output: Applies a verification protocol to eliminate false positives, then writes severity-calibrated findings (Critical, Major, Minor, Informational) with confidence levels to an output file. - Use Case: When a teammate opens a PR adding a new skill to your marketplace, run this review to automatically flag an invalid frontmatter, a name collision with an existing skill, or a vague description before merge. ## Quick Start Review the skill changes in this PR against the main branch and write the findings to .review-output.md.
