Audits whether your project docs actually stay on topic and cover what they claim to. Checks each file's stated SCOPE against its content, flags coverage gaps and scope creep, and watches for SSOT violations across your documentation set. Targets the usual suspects like architecture.md, tech_stack.md, and runbook files. It's opinionated about agent instruction files too, warning when AGENTS.md bloats with linter rules that belong in Biome or conditional patterns that should use Claude's path scoping instead. Scores based on coverage, relevance, and focus, then writes structured findings with line numbers and fix suggestions. Not a fact checker, just a semantic fitness judge that keeps docs aligned with their declared purpose.
npx skills add https://github.com/levnikolaevich/claude-code-skills --skill ln-612-semantic-content-auditor