From mindmap
Core rules for mind map generation including stable IDs, evidence requirements, mandatory publish step, artifact store flow, and hierarchical parsing strategy.
How this skill is triggered — by the user, by Claude, or both
Slash command
/mindmap:mindmap-rulesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- File IDs are deterministic SHA-256 hashes of canonical paths
discover stores DiscoveryResult → returns artifactIdinspect(parse) stores InspectionResult + ParseResult → returns an outer artifactId (inspectionResult) AND an inner parseArtifactIdgenerate reads DiscoveryResult + ParseResult, builds MindMapJSON, validates, and publishesAlways pass the outer artifactId from inspect downstream — both as inspectionArtifactId on mindmap.generate and as appendToInspectionArtifactId on subsequent inspect calls. Never pass the inner parseArtifactId; generate looks it up internally. Passing the inner ID produces "inspection artifact … missing or not parse-based" errors.
mindmap.generate only writes to disk if options.writePath is provided. Without it, the published mind map is returned via the MCP structuredContent channel but mindmap-output.json will NOT be updated. Always pass writePath pointing at the target repo's mindmap-output.json.
npx claudepluginhub engincankaya/claude-code-mind-map --plugin mindmapCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.