| name | graph |
| description | Knowledge graph analysis -- statistics, clusters, paths, bridges, PageRank |
| version | 3.0.0 |
| triggers | ["/graph"] |
| type | skill |
| tags | ["skill","agent","discovery","graph","analysis"] |
Graph -- Knowledge Graph Analysis
What it does
Builds a directed graph from vault notes and their wiki-links, then runs graph algorithms to reveal structure, importance, and hidden patterns.
Usage
/graph → Full graph statistics
/graph clusters → Topic clusters by shared tags
/graph rank → Top notes by PageRank importance
/graph path "Note A" "Note B" → Shortest path between two notes
/graph bridges → Notes that connect otherwise separate groups
/graph similar "Note" → Notes with similar connection patterns
/graph hops "Note" → Hidden connections 2-3 hops away
How it works
- Calls
ensureFreshIndexes('.') from scan utils (rebuild if stale)
- Loads vault-index.json, link-map.json, tag-index.json
- Calls
loadAndBuild('.') from graph-engine.cjs to construct the graph
- Runs the requested analysis algorithm
- Formats and presents results
Algorithms
| Command | Algorithm | What it reveals |
|---|
| (default) | graphStats() | Node/edge count, density, components, degree distribution, orphans |
| clusters | findTagClusters() | Groups of 3+ notes sharing the same tag |
| rank | pageRank() | Notes ranked by structural importance (link authority) |
| path | shortestPath() | Minimum hops between any two notes |
| bridges | findBridgeNotes() | Notes whose removal would split the graph |
| similar | findStructurallySimilar() | Notes linking to/from the same neighbors (Jaccard similarity) |
| hops | findMultiHopConnections() | Notes reachable in 2-3 hops but not directly linked |
Dependencies
const { loadAndBuild, graphStats, pageRank, findTagClusters, shortestPath,
findBridgeNotes, findStructurallySimilar, findMultiHopConnections
} = require('./.agents/skills/graph/graph-engine.cjs');
const { ensureFreshIndexes } = require('./.agents/skills/connect/connect-utils.cjs');
Output Format
Results are presented as formatted text tables. Example:
Knowledge Graph | 57 nodes, 238 edges
Density: 0.074 (sparse -- room to grow)
Components: 2 (1 main + 1 isolated)
Avg degree: 8.35 links per note
Top 5 by PageRank:
1. Home (0.082)
2. Universal AI Clothing Kit (0.041)
3. Projects MOC (0.038)
...
Governance
- AUTO: Read indexes, compute graph metrics (read-only, no vault changes)
- All graph operations are non-destructive
- Never modifies files -- only analyzes and reports
