Browse Skills

Docs Corpus Audit runs a whole-corpus regrounding pass over the documentation against current source code, page by page, and emits surgical fixes for stale claims. Use it when maintainers want a full docs audit, not per-PR drift review.

Docs Grounding Verifier verifies claim-level grounding of documentation pages against source code, returning per-claim verdicts with file:line evidence citations. Use it when you need to fact-check specific pages for accuracy, not for whole-corpus audits or PR triage.

Python Architecture guides the creation of new modules, refactoring class hierarchies, introducing design patterns, and making multi-file changes in the APM CLI codebase. Use it to follow existing patterns and centralize shared logic with regression tests and static checks.

Validate Notes checks a note's YAML frontmatter, required fields, enum values, description quality, and wiki-link health against domain-specific templates. It warns rather than blocks capture, and triggers on '/validate', '/validate [note]', 'check schema', 'validate note', or 'validate all'.

Vault Health runs condition-based vault diagnostics across schema, orphans, links, description quality, three-space boundaries, throughput, stale notes, and MOC coherence. It supports quick, full, and three-space modes and returns actionable FAIL/WARN/PASS reports.

Verify Notes checks note descriptions via cold-read prediction, validates schema compliance, verifies wiki links, and confirms topic map integration. Use it as a quality gate after creating notes or as periodic maintenance.

Review a Design evaluates whether a design is sound—solving the right problem and following from its stated goals and constraints—and returns ranked, evidence-backed findings, not edits. Use it to pressure-test proposals, specs, ADRs, or architecture decisions before committing to them.

Check Citations verifies academic citations against CrossRef, Semantic Scholar, and OpenAlex to detect AI-hallucinated references, chimeric citations, and suspicious patterns. Use it before submitting papers or when auditing bibliographies.

Code Health scans the codebase for dead code, tech debt, outdated dependencies, and code quality issues using the Centinela (QA) agent. Use it before releases, after refactoring, or for periodic codebase hygiene checks.

Release Check performs a pre-release verification checklist, validating features, tests, documentation, security, and quality gates before shipping. Use it before production releases or version tags as a final quality gate.

Review Findings addresses and fixes findings from a QA code review. It reads the review report, fixes critical and warning issues, updates tests, and prepares a fix report for re-verification.

Security Audit performs a deep security review covering OWASP Top 10, authentication, authorization, data protection, dependency vulnerabilities, and secrets scanning. Use it before releases or after security-sensitive changes to produce a findings report.