Long-term memory for AI assistants. Graph + vector store that recalls decisions, relationships, and context across sessions.
Terminal score: 0–1000 raw, weighted across 4 dimensions. Public score: 0–10 normalized (shown in the 30-day stars chart above).
