dual-axis-skill-reviewer

Automate AI skill quality checks by integrating a dual-axis reviewer script into your CI/CD pipeline to automatically score new skills, gate merges based on thresholds like 90+, and generate actionable feedback.

Dual-axis LLM evaluation combines deterministic code-based checks verifying skill structure and script execution safety with LLM-driven deep content analysis of logic, correctness, and maintainability to produce a reproducible quality score.

Validate AI skill directory structure and test coverage by running an automated reviewer script that checks skill structure, script executability, and test coverage, saving reports to a specified reports directory.

You can analyze AI skills across different project roots by specifying a project root parameter when running the dual-axis reviewer, enabling cross-project skill structure validation and LLM review findings.

You need pyyaml installed as a required dependency to run the automated skill review scripts that perform deterministic code-based checks and generate reproducible quality scores.

LLM-driven deep content analysis evaluates logic, correctness, and maintainability beyond deterministic code checks, ensuring AI skills meet qualitative assessments for reliable and safe operation.