Show HN: Persistent Memory for Claude Code (MCP) (github.com)

A-MEM, a self-evolving memory system for coding agents, has been introduced and tested with Claude Code, aiming to enhance knowledge management in AI environments. Unlike traditional vector stores, A-MEM organizes memories into a dynamic Zettelkasten-style graph that evolves over time as new information is added. This innovative system not only allows memories to be stored but also facilitates their growth and connection through semantic and structural searches. It automatically extracts keywords and context, ensuring that knowledge remains relevant and interconnected. This development is significant for the AI/ML community as it addresses the challenge of static knowledge management by creating a more intelligent and adaptive system. By employing a combination of semantic similarity and graph traversal, A-MEM allows agents to not just retrieve information, but also understand and utilize relationships between memories more effectively. With six accessible tools for managing memory—such as adding, searching, and updating notes—A-MEM offers an efficient framework for coding agents to leverage stored knowledge, ultimately improving their learning capabilities and adaptability in real-time applications.