| namespace | aiwg |
| platforms | ["all"] |
| name | profile-communities |
| description | Detect co-author communities (label propagation), report modularity Q, and identify bridge authors (high betweenness). Runs via `aiwg corpus profile-communities`. |
| commandHint | {"argumentHint":"[--out PATH]","allowedTools":"Read, Bash","model":"haiku","category":"research-analytics","modelRole":"efficiency","modelTier":"economy"} |
Profile Communities
Community structure of the corpus co-author graph.
How to run
aiwg corpus profile-communities
aiwg corpus profile-communities --out indices/communities.md
Uses label propagation (deterministic variant — no python-louvain
dependency; the graceful fallback the corpus tooling specifies), reports
modularity Q (warns when Q < 0.3 = weak structure on a sparse graph), and
lists bridge authors (highest betweenness, spanning communities).
Triggers
- "detect research communities"
- "co-author clusters / modularity"
- "bridge authors"
/profile-communities
References
- @$AIWG_ROOT/src/artifacts/corpus-tools/profile-communities.ts — implementation
- profile-metrics / profile-temporal skills
