Help us improve
Share bugs, ideas, or general feedback.
From science-agent
Verifies [NB##:K##] claim references in prose against notebook sources, surfacing unresolved or stale numeric references.
npx claudepluginhub andyed/science-agent --plugin science-agentHow this command is triggered — by the user, by Claude, or both
Slash command
/science-agent:notebook-auditThe summary Claude sees in its command listing — used to decide when to auto-load this command
Run notebook-audit with the user's arguments: Surface unresolved claim references and stale numerics. If the project hasn't set up Key Claims conventions, the tool degrades gracefully and explains what to do — pass that explanation through.
/askAsks NotebookLM a question about the current notebook's content via `notebooklm ask` CLI, displaying the response with optional source references via --json.
/verify-citationsVerifies citations in markdown files against research corpus, detects hallucinations and GRADE violations, generates report with issues, PASS/FAIL verdict, and supports auto-fix.
/claude-md-validateValidates CLAUDE.md files across directory tree for references, structure, consistency, content quality, integration, and deployment readiness, generating JSON reports and markdown summaries.
/healthRuns unified documentation health analysis detecting stale docs, broken links, version drift, and cross-document inconsistencies. Supports modes like scan, links, check <path>, drift, and setup.
/fact-checkVerifies factual claims in a document (HTML/MD or recent diagram) against codebase and git history, corrects inaccuracies inline, adds verification summary.
/checkFact-check content against the Midnight ecosystem. Extracts claims, classifies by domain, verifies each via midnight-verify, and produces a report.
Share bugs, ideas, or general feedback.
Run notebook-audit with the user's arguments:
NODE_PATH="${CLAUDE_PLUGIN_DATA}/node_modules" node "${CLAUDE_PLUGIN_ROOT}/cli.js" notebook-audit $ARGUMENTS
Surface unresolved claim references and stale numerics. If the project hasn't set up Key Claims conventions, the tool degrades gracefully and explains what to do — pass that explanation through.