10x-rule-review

Reviews AI rules files and produces a five-check scorecard with concrete fixes.

Updated May 23, 2026
One-click install
npx skills add https://github.com/BMS-kmusial/bms-stream-monitor --skill 10x-rule-review-bms-kmusial
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: 10x-rule-review
Source: https://github.com/BMS-kmusial/bms-stream-monitor/tree/main/.claude/skills/10x-rule-review
Command: npx skills add https://github.com/BMS-kmusial/bms-stream-monitor --skill 10x-rule-review-bms-kmusial

SYSTEM DOCUMENTATION & REQUIREMENTS

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 rule-for-AI markdown file and returns a structured scorecard with actionable fixes. ## Core Features & Use Cases - Five-dimension audit: Scores length, direct code snippets, vague language, redundant knowledge, and rule ordering with OK/WARN/FAIL verdicts. - Grounded rewrites: Translates vague phrases like "write clean code" into testable rules based on the project's actual stack and conventions. - Interactive reordering: Proposes a better section order based on U-shaped attention and asks for approval before moving any content. - Use Case: Point it at your repo's AGENTS.md after months of accretion to find duplicated README content, buried security rules, and snippets that should be @-references instead. ## Quick Start Ask the assistant to run /10x-rule-review on your CLAUDE.md or AGENTS.md file to get a scored report with prioritized fixes.

Frequently Asked Questions about 10x-rule-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review my CLAUDE.md or AGENTS.md file quality?

Invoke /10x-rule-review with the path to your rules file, such as @CLAUDE.md or AGENTS.md. The skill reads the file, runs five checks covering length, snippets, language precision, redundancy, and ordering, then prints a scorecard with path:line references for every finding.

What makes a good AI rules file for coding agents?

A healthy rules file stays under 200 non-blank lines, avoids embedded code snippets in favor of @-references, uses testable rules instead of vague phrases, omits knowledge the model already has, and places critical rules at the top. This skill scores each of these dimensions explicitly.

Does the rule review work with .cursor/rules .mdc files?

Yes, the skill is tool-agnostic and treats .mdc files, .windsurfrules, copilot-instructions.md, and nested per-area rule files the same way. For .mdc files, frontmatter like globs and alwaysApply is excluded from line counts since it is configuration, not rule content.

Will the review edit or rewrite my rules file automatically?

No, the default output is read-only. The only possible edit is a section reorder from Check 5, and that happens only after you explicitly approve it through an interactive prompt. Rule content is never rewritten, only moved.

Why does my rules file get flagged for redundant knowledge?

Paragraphs restating framework defaults, lint-enforced rules, generic definitions, or README content are flagged because the model already knows them and they waste context. The skill suggests deleting them, replacing them with @-references, or keeping them only if tied to a documented incident.