dual-axis-skill-reviewer

Automate code quality scoring for skills by running the dual-axis reviewer script with `--project-root` to apply deterministic checks and LLM deep review across projects, generating reproducible scores in the reports directory.

Dual-axis skill review combines deterministic structural checks for scripts, tests, and execution safety with LLM-based deep code review findings to produce reproducible quality scores and concrete improvement items for SKILL.md files.

Gatekeep skill merges by running the dual-axis reviewer to generate reproducible quality scores, then enforcing a score threshold like 90+ before allowing merges, supported by concrete improvement items for low-scoring skills.

Yes, automated skill reviews require Python 3.9+ and the uv toolchain for execution and reporting, utilizing dependencies such as pyyaml to process and evaluate SKILL.md files.

Generate actionable improvement items for low-scoring skills by executing the dual-axis review script, which merges deterministic structural checks with LLM deep review findings to output concrete remediation steps.

Yes, you can review SKILL.md files in repositories outside the current directory by passing the target path to the `--project-root` argument, enabling cross-project skill review and reporting.