DeepMemory
Provides long-term memory storage for conversational agents, accessible via a Model Context Protocol server.
Acerca de
DeepMemory is a Model Context Protocol (MCP) server designed to equip conversational AI agents with robust long-term memory capabilities. It stores memories locally in an SQLite database by default, offering a comprehensive set of MCP tools for adding, searching, listing, updating, and deleting these memories. For scalable solutions, it also supports an optional MySQL backend. The server features 'Clusters' to organize structured details and link related memories, a persistent queue to handle memory write requests during initialization, and a lightweight HTTP fallback for basic interactions. Additionally, it provides dedicated storage and tools for managing development-related documents, helping to centralize knowledge for AI projects.
Características Principales
- Comprehensive MCP tools for memory lifecycle management (add, search, update, delete)
- Clustering mechanism to group structured details and link associated memories
- Flexible data storage with default SQLite and optional MySQL backend support
- Persistent write queue to ensure no memory writes are lost during server initialization
- Dedicated storage and tools for managing development-related documents (docs)
- Lightweight HTTP fallback for health checks and minimal interactions
- 0 GitHub stars
Casos de Uso
- Integrating a memory service with MCP host environments like Jan for enhanced AI capabilities
- Providing long-term memory and knowledge base for AI conversational agents
- Organizing and retrieving structured information related to AI agent's topics or projects