/hotspots

/wicked-garden:search:hotspots

Identify hotspot symbols ranked by total reference count (incoming + outgoing). Use this to find central coupling points, high-impact refactor targets, and architectural load-bearers.

Instructions

1. Search brain for high-connectivity symbols:

   curl -s -X POST http://localhost:4242/api \
     -H "Content-Type: application/json" \
     -d '{"action":"search","params":{"query":"class function module export","limit":50}}'
   

   If brain is unavailable, fall back to Grep/Glob:

   • Use Grep to find class/function definitions across the codebase
   • Use Grep to count import/require references to each symbol
   • Rank by reference count Suggest wicked-brain:ingest to index the codebase for faster hotspot analysis.

2. Count references for discovered symbols using Grep:

   Grep: <symbol_name> across all source files — count matches
   

3. Arguments (all optional):

   Argument	Default	Description
   --limit N	20	Number of results to return
   --layer LAYER	—	Filter to one architectural layer (backend, frontend, database, view)
   --type TYPE	—	Filter to one symbol type (CLASS, FUNCTION, METHOD, TABLE, etc.)

4. Present results as a ranked table:

   Rank	Symbol	Type	Layer	In	Out	Total
   1	UserService	CLASS	backend	42	18	60
   2	db_session	FUNCTION	database	38	5	43
   3	AuthMiddleware	CLASS	backend	31	11	42

5. Highlight:

   • The top 3-5 symbols and why they are hotspots (high in-count = shared dependency; high out-count = coordinator/orchestrator)
   • Any layer with a disproportionate share of hotspots (potential coupling concentration)
   • Symbols where in_count is very high relative to out_count (widely imported utilities)
   • Symbols where out_count is very high relative to in_count (orchestrators or god objects worth reviewing)

Example

/wicked-garden:search:hotspots
/wicked-garden:search:hotspots --limit 10
/wicked-garden:search:hotspots --layer backend
/wicked-garden:search:hotspots --type CLASS --limit 15