What problem does it solve? AI rules files like CLAUDE.md, AGENTS.md, or .cursor/rules/*.mdc often grow bloated, vague, and poorly ordered, which wastes context window space and causes agents to ignore critical instructions. This Skill audits any rules-for-AI markdown file and tells you exactly what to fix. ## Core Features & Use Cases - Five-Check Scorecard: Scores the file on length, direct code snippets, vague language, redundant knowledge, and rule ordering with OK/WARN/FAIL verdicts. - Grounded Fixes: Translates vague phrases like "write clean code" into testable rules based on your project's actual stack and conventions. - Interactive Reordering: Detects buried critical rules and, with your explicit approval, reorders sections so high-priority rules get the most model attention. - Use Case: Run it on your repo's AGENTS.md before onboarding a new AI coding assistant to find redundant paragraphs, replace duplicated docs with @-references, and surface security rules buried at the bottom of the file. ## Quick Start Ask the assistant to review your AI rules file by passing its path, for example: review my AGENTS.md file and score it.
