agentmem

agentmem is a lightweight Model Context Protocol (MCP) server that provides persistent, searchable memory for AI agents. It exposes 16 distinct memory management and context extraction tools via an HTTP REST API and direct MCP server mode. Designed for developers, AI platform builders, and advanced AI workflows, agentmem allows agents to store, recall, consolidate, and extract structured knowledge from chats, documents, and task histories—surviving across sessions via a single SQLite file. It includes hybrid search (keyword + semantic), entity extraction (NER), namespace isolation for multi-agent scenarios, and procedural/episodic memory organization suitable for robust persistent context in any AI-powered environment.