This server addresses the common limitation of AI assistants forgetting information between sessions by providing persistent memory capabilities. It integrates with Velixar's memory API, exposing a suite of five tools that allow any Model Context Protocol (MCP)-compatible AI client—such as Claude Desktop, Kiro, Cursor, or Windsurf—to store, search, list, update, and delete memories. These memories persist across conversations, restarts, and even different machines, ensuring AI assistants retain crucial context and knowledge for improved, ongoing interactions.
主な機能
010 GitHub stars
02Browse and paginate through existing memories
03Store memories with optional tags and memory tiers
04Update the content of existing memories
05Perform semantic searches across stored memories
06Delete specific memories by ID
ユースケース
01Maintaining current conversation context beyond single interactions
02Endowing AI assistants with long-term recall across various conversations and sessions
03Storing critical facts or general long-term knowledge for AI agents
04Facilitating shared organizational knowledge access for AI assistants
