Manages agentic memory for LLM agents, offering direct integration into IDEs like Cursor and VSCode via the MCP protocol.
Sponsored
A-MEM is an innovative agentic memory system designed to empower LLM agents by leveraging the Zettelkasten principle. It distinguishes itself through direct integration with popular IDEs such as Cursor and VSCode via the MCP protocol, providing developers with seamless access to advanced memory functionalities. Its unique dual-storage architecture combines ChromaDB for vector similarity searches with NetworkX for explicit, typed graph-based memory relationships, enabling intelligent semantic retrieval and dynamic memory evolution directly within your development workflow.
主な機能
01Dual-storage architecture using ChromaDB for vector search and NetworkX for explicit typed graph relationships
02Intelligent semantic retrieval leveraging graph traversal for related memories
03Automatic note construction with keyword, tag, and summary extraction
04Direct IDE integration via MCP protocol for Cursor and VSCode
050 GitHub stars
06Dynamic memory evolution and automatic linking of similar memories
ユースケース
01Enhancing LLM agent capabilities with persistent, structured memory within IDE environments
02Integrating advanced memory management directly into developer workflows for AI-assisted coding
03Developing and testing LLM agents that require Zettelkasten-inspired information recall and evolution
