Instruction QA Audit
Performs static-analysis QA passes on AI instruction files and configuration prose to eliminate contradictions, dead logic, and unstated dependencies.
Instruction QA Audit conducts comprehensive static-analysis reviews of instruction and configuration prose, including SKILL.md files, CLAUDE.md files, global context files, and slash command templates, without touching application source code. By dispatching dedicated Opus subagents to analyze each target document in parallel, the skill identifies internal contradictions, unstated dependencies, dead instructions, and frontmatter guideline violations. It then compiles findings ranked by severity, proposes clear remediations, and waits for explicit user confirmation before applying targeted updates, ensuring your AI operational guidelines remain clean, aligned, and effective.