Long-Term Memory

Memory MCP Server provides long-term memory storage for LLMs using SQLite and vector embeddings. It enables AI assistants to store, retrieve, and search through different types of memories (conversations, code, decisions, references) with semantic search capabilities powered by Ollama's nomic-embed-text model. The implementation includes a project-based organization system with tagging support, relationship tracking between memories, and a set of MCP tools for memory interaction. This server is particularly valuable for maintaining context across multiple conversations, preserving development decisions, and building knowledge bases that persist beyond individual chat sessions.