Skip to main content

Detect Skills

Evidence-based repo audits with confidence-scored findings.

Pipeline​

Run individual detect skills for specific domains, or run all 5 then consolidate:

/jaan-to:detect-dev        → docs/current/dev/
/jaan-to:detect-design     → docs/current/design/
/jaan-to:detect-writing    → docs/current/writing/
/jaan-to:detect-product    → docs/current/product/
/jaan-to:detect-ux         → docs/current/ux/
                                    ↓
/jaan-to:pack-detect       → docs/current/{README,risk-heatmap,unknowns-backlog,source-map}.md

Available Skills​

SkillDescriptionOutput Files
/jaan-to:detect-devEngineering audit with OpenSSF-style scoring9 files
/jaan-to:detect-designDesign system detection with drift findings6 files
/jaan-to:detect-writingWriting system extraction with NNg tone scoring6 files
/jaan-to:detect-productProduct reality extraction with 3-layer evidence7 files
/jaan-to:detect-uxUX audit with Nielsen heuristics and journey mapping7 files
/jaan-to:pack-detectConsolidate all detect outputs into scored index4 files

Shared Standards​

All detect skills share:

  • Evidence format: SARIF-compatible blocks with id, type, confidence, location (uri, startLine, snippet), method
  • Evidence ID namespaces: E-DEV-NNN, E-DSN-NNN, E-WRT-NNN, E-PRD-NNN, E-UX-NNN (prevents collisions in pack-detect)
  • 4-level confidence: Confirmed (0.95–1.00) / Firm (0.80–0.94) / Tentative (0.50–0.79) / Uncertain (0.20–0.49)
  • Universal frontmatter: title, id, version, status, target.commit, tool metadata, confidence_scheme, findings_summary, overall_score (0–10), lifecycle_phase
  • Document structure (Diataxis): Executive Summary → Scope/Methodology → Findings → Recommendations → Appendices
  • Output path: docs/current/{domain}/ in the target project (NOT $JAAN_OUTPUTS_DIR)

Reference​

Back to Skills