skill-improver

Automate quality improvements for Claude Code skills using the skill-reviewer agent from the plugin-dev ecosystem. It runs iterative review cycles that validate frontmatter, categorize issues into critical, major, and minor, and apply fixes until standards are met.

Yes, you can enforce standardization and frontmatter validation in Claude Code skills. The automated improvement loop validates frontmatter and applies policy-driven governance to ensure skills meet quality standards during the fix-review cycle.

Categorize and fix quality issues in Claude Code agents by running automated reviews that classify problems into critical, major, and minor severity levels. The structured iteration loop then applies targeted fixes and re-evaluates the skill.

Yes, the plugin-dev ecosystem is required for automated skill reviews. The skill-improvement process specifically requires the skill-reviewer agent provided through this ecosystem to execute the structured fix-review cycles.

The best way to iteratively refine Claude Code skills with multiple quality issues is to use an automated fix-review loop. This process repeatedly surfaces issues, applies fixes, and re-evaluates until the skill meets all standardization requirements.