Evaluate rule files and bootstrap documents with rubric-driven scoring.
npx skills add https://github.com/sfc-gh-myoung/ai_coding_rules --skill rule-reviewer
Please help me install this Agent Skill. Skill: rule-reviewer Source: https://github.com/sfc-gh-myoung/ai_coding_rules/tree/main/skills/rule-reviewer Command: npx skills add https://github.com/sfc-gh-myoung/ai_coding_rules --skill rule-reviewer
The Rule Reviewer provides deterministic, rubric-based evaluation of autonomous-rule content to ensure quality, completeness, and cross-agent compatibility, surfacing actionable gaps and standardizing reviews across agents.
Load the Rule Reviewer and run a FULL review on rules/200-python-core.md with model claude-sonnet-45 on 2026-01-06.
High-intent search queries and answers about installing and using this skill.
Perform a rubric-driven rule review by loading the evaluator on your rules files to receive deterministic scoring across six dimensions and actionable quality signals. The process includes input validation, schema checks, and evidence-backed recommendations.
Cross-agent compatibility for rule content ensures autonomous AI guidance remains parsable, consistent, and actionable across different agents. The review applies a deterministic rubric to surface gaps and standardize quality assurance across agents.
Yes, rubric scoring supports project bootstrap files like AGENTS.md and PROJECT.md alongside standard rule files. The evaluator applies schema-aware markdown checks and explicit input validation to ensure bootstrap documents meet quality standards.
Yes, the rule review process writes evaluations to a dedicated reviews directory with no-overwrite safety and automatic suffixing when needed. This ensures existing review files remain intact while new evaluations are saved safely.
Deterministic rule evaluations focus on six rubric dimensions including staleness, parsability, and token efficiency but require properly formatted markdown rule files or bootstrap documents. They rely on schema-aware checks and explicit input validation to generate evidence-backed recommendations.
lukasPWRAnalyze AI rule files and generate a 5-point scorecard.
harkersAssess vendor privacy and security posture with evidence-backed findings.
jayhjenkinsAutomate three-tier QA testing with TEST_INDEX.md and RUNS/ reports.
pproencaValidate that every rule has a corresponding Markdown reference file.
