second-brain-graph

1# Analyse the graph
2 
3The graph is the search space every answer is drawn from. Its shape decides
4whether retrieval reaches the relevant material in a few hops or reaches nothing.
5 
6## Core rule
7 
8Report shape and what it means, never fix anything. This skill is read-only.
9 
10## Workflow
11 
121. **Run the scripts** rather than reading files: `scripts/vault_stats.py` for
13   counts and degree, `scripts/link_check.py` for orphans and breaks,
14   `scripts/graph_export.py` when deeper analysis is wanted.
152. **Compute the four metrics:** orphan rate, average degree, component count,
16   stale-page rate.
173. **Identify hubs** (pages that have swallowed the graph) and **bridges**
18   (pages connecting otherwise separate clusters).
194. **Name the clusters.** These are what the vault is actually about, which is
20   reliably different from what the user would say it is about.
215. **Say what each number means** for answering questions, not just the value.
22 
23## Output format
24 
25```
26Pages <n> | links <n> | avg degree <n>
27Orphan rate <n>%      (healthy under 5)
28Components <n>        (main holds <n>%)
29Stale concepts <n>%   (untouched 90+ days)
30 
31Hubs: <pages with outsized degree, and whether to split them>
32Bridges: <pages connecting clusters, and what breaks if they are wrong>
33Clusters: <what the vault is actually about>
34 
35What this means: <two or three sentences>
36```
37 
38## Calibration
39 
40Never report page count or word count as a health signal. Both rise whether the
41vault is improving or not.
42…