exploring-graph

npx skills add majiayu000/claude-skill-registry --skill exploring-graph-alexanderop-second-brain-nuxt --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

# Exploring the Knowledge Graph This skill analyzes the structure of connections in the knowledge base to surface insights. ## Workflow ### 1. Build the Connection Map Parse all wiki-links from every content file: ```bash # Extract all wiki-links grep -oh '\[\[[^]]*\]\]' content/*.md # Get links per file for f in content/*.md; do echo "=== $f ===" grep -o '\[\[[^]]*\]\]' "$f" done ``` Build an adjacency list: - `note-a` → links to: `[note-b, note-c]` - `note-b` → links to: `[note-a]` - etc. ### 2. Calculate Metrics **For each note, determine:** | Metric | Description | |--------|-------------| | Outgoing links | Wiki-links in this note | | Incoming links | Other notes linking to this one | | Total connections | Outgoing + Incoming | ### 3. Identify Patterns #### Hub Notes (Most Connected) Notes with the highest total connections - these are central to the knowledge base. #### Orphan Notes - **Full orphans**: 0 incoming AND 0 outgoing links - **Dead ends**: Has outgoing but 0 incoming - **Sources**: Has incoming but 0 outgoing #### Clusters Groups of notes that link heavily to each other but less to the rest. Identify by: - Shared tags - Mutual links - **Map membership** - notes lin

Extract all wiki-links
grep -oh '\[\[[^]]*\]\]' content/*.md
Get links per file
for f in content/*.md; do
echo "=== $f ==="
grep -o '\[\[[^]]*\]\]' "$f"
done
grep -l "type: map" content/*.md