Skill Dependency Graph

Contract

• Input: problem or task defined by the skill body.
• Output: Markdown artifact or structured result with completion criteria met.
• Side effects: none (design/review/documentation only unless explicitly stated).
• Dependencies: none (self-contained unless linked to other skills).
• Stop condition: all process steps completed; artifact saved; criteria checked.
• Risk: low.
• Boundary: produces reasoning or documentation artifacts; does not modify external systems unless explicitly instructed.

Skill Dependency Graph

Contract

• Input: the canonical Skills directory.
• Output: Mermaid or JSON graph, cycle list, central Skills, and modularity findings.
• Boundary: analyze declared dependencies only; do not rewrite Skill contracts.

Run node .agents/skills/platform/skill-lab.mjs graph --format mermaid for a diagram or --format json for analysis input. Use incoming-edge counts to identify central Skills and inspect every cycle before adding another dependency.