Local LTM

概要

Local LTM is a modular, locally-maintained Long-Term Memory system designed to provide AI agents with durable memory capabilities. Operating as a Model Context Protocol (MCP) server, it allows a single user to run multiple AI agents that can store, search, update, and delete persistent knowledge, preferences, facts, and skills. It leverages Mem0 as its memory backend, offering structured storage for episodic, semantic, and procedural memory types, ensuring AI agents can maintain context and learn across conversations.

主な機能

• Locally-maintained long-term memory system for AI agents
• Store, search, update, and delete durable memories (CRUD operations)
• Integrates with Mem0 as the memory backend
• Exposed via Model Context Protocol (MCP)
• 2 GitHub stars
• Supports episodic, semantic, and procedural memory types

ユースケース

• Storing user preferences and facts for AI agents
• Managing agent-specific heuristics and rules across sessions
• Providing persistent knowledge for AI agents to reference across conversations