Audit review skills in .claude/skills for broken references and inconsistencies.
npx skills add https://github.com/cwilliams5/Alt-Tabby --skill review-reviews
Please help me install this Agent Skill. Skill: review-reviews Source: https://github.com/cwilliams5/Alt-Tabby/tree/main/.claude/skills/review-reviews Command: npx skills add https://github.com/cwilliams5/Alt-Tabby --skill review-reviews
This Skill ensures that all other review skills in the codebase are accurate, up-to-date, and consistent with the project's architecture and rules, preventing outdated or misleading information from causing errors.
review-reviews to catch any broken references or outdated instructions that could lead the agent astray.Run the review-reviews skill to audit all skills in the .claude/skills directory for mechanical inaccuracies.
High-intent search queries and answers about installing and using this skill.
To audit code review skills, run the review-reviews skill to check all skills in the .claude/skills directory for mechanical inaccuracies, stale references, and drifted architecture claims. It classifies findings as Broken, Renamed, or Current to ensure codebase consistency.
Skill consistency validation verifies file paths, function names, tool flags, and configuration keys against the current codebase. You need it before deploying AI agents that rely on review skills, preventing outdated instructions from causing agent validation errors.
Verifying tool flags and IPC message types requires cross-referencing skill documentation against actual tool implementations. The review-reviews skill checks if parameters and usage patterns described in skills still match the current codebase and identifies tool flag mismatches.
Yes, a skill audit catches architecture contradictions by verifying skills against canonical architecture and coding rules. It identifies cross-skill inconsistencies and drifted architecture claims, ensuring all review skills remain accurate and aligned with the project's current rules.
The best way to check skill cross-references is running an identifier cross-reference audit across the .claude/skills directory. This verifies file paths, function names, and skill cross-references against the current codebase to catch broken references that could lead an agent astray.
enbiyagoralGenerate a standardized SKILL.md and directory scaffold for Claude Code skills.
joshua3403Analyze a codebase for architecture consistency and convention adherence.
dorrioExtract claims from Markdown files and verify them against the codebase.
l3ochoIdentify and flag drift between documentation references and the codebase.