The problem this solves
Agents need to follow relationships: which preference belongs to which project, which decision came from which incident, and which insight should be reused by other agents.
Feature / Open Source AI Memory Database
Knowledge Graph Memory for Agents
Memorose uses graph-linked memory to connect facts, procedures, insights, and shared organization knowledge instead of treating every memory as an isolated vector.
How Memorose handles it
Typed memory relationships and graph-aware retrieval let agents expand context when relationships matter, while keeping scope controls visible.
runtime example
curl -X POST "$BASE_URL/v1/users/dylan/streams/$STREAM_ID/retrieve" \
-H "Authorization: Bearer $TOKEN" \
-H "Content-Type: application/json" \
-d '{"query":"related deployment decisions","graph_depth":2}'