GraphMemory-IDE
An MCP server that gives AI coding assistants long-term, on-device memory inside an IDE, backed by the Kuzu graph database with semantic vector search, and integrated with VS Code, Cursor, and Windsurf.
GraphMemory-IDE is an MCP (Model Context Protocol) server that gives AI-assisted development tools a long-term memory system, combining graph-based knowledge storage in the Kuzu native graph database with semantic vector search over HNSW indexes and sentence-transformer embeddings, plus real-time analytics streamed over WebSocket and SSE. Dedicated plugins integrate it with VS Code, Cursor, and Windsurf.
Graph algorithms such as centrality, community detection, path analysis, and similarity computation can optionally compile through Codon for a reported 10-100x speedup, falling back to Python and NetworkX automatically when Codon is unavailable. The backend is FastAPI with JWT (EdDSA/Ed25519) authentication and rate limiting, with a Streamlit dashboard for graph metrics, user activity, and system health, and Prometheus/Grafana for observability.
A Docker Compose deployment runs the full stack, including Nginx, PostgreSQL, Redis, and monitoring, alongside a lighter local-development path with uvicorn and streamlit run.