agentic-os/brain/graph/README.md

1.7 KiB

Central Brain — Graphify knowledge graph

One central, queryable brain that project graphs roll up into.

Topology

  • agentic-os.json — this project's own code graph (built from graphify . --code-only).
  • projects/<name>.json — each linked project's graph (registered via link-project.sh).
  • central-graph.json — the MERGED cross-repo brain (union of all the above). This is the artifact everything queries.
  • serve-brain.sh — serves central-graph.json over HTTP MCP at http://127.0.0.1:8090/mcp.
  • Dashboard: the brain directory is mounted at http://127.0.0.1:8080/brain-graph/.

Commands

bash brain/graph/build-central.sh            # rebuild central brain (code graph + merge linked projects)
bash brain/graph/link-project.sh <name> <graph.json>   # register a project's graph into the brain
bash brain/graph/serve-brain.sh             # serve central brain over HTTP MCP (:8090)

Set CENTRAL_BRAIN_PORT to override the serve port (default 8090).

What it contains today

  • Code symbols of agentic-os as nodes (functions/classes/modules), with call/import edges.
  • Linked projects (added via link-project.sh).
  • Memory NOTES (markdown under brain/) are NOT yet graph nodes. To make them queryable, set an LLM key (GEMINI_API_KEY / OPENAI_API_KEY / etc.) and run graphify . (no --code-only) so notes are semantically extracted. Until then, notes remain the write path; the graph indexes code structure.

Notes

  • graphify's MCP HTTP serve requires the extra: pip install "graphifyy[mcp]" (done in the venv).
  • The write path for memory is still brain/*.md + audit + data/; graphify reads those, never writes them.
  • Rebuild is scheduled hourly by the agentic-os-brain-rebuild cron.