tverney/mcp-agent-memory

What it does

tverney/mcp-agent-memory is an MCP (Model Context Protocol) server that extends AI agents with persistent memory functionality. It allows agents to store, retrieve, and search information across sessions through a filesystem-based storage system. The tool implements intelligent memory consolidation through a native daemon process that automatically processes and extracts insights from stored memories.

How to set up

Installation is straightforward using npm: run npx mcpagentmemory setup to initialize the server. The tool is compatible with multiple MCP clients including Kiro, Claude Desktop, Cursor, and other standard MCP-compatible applications. Once configured, agents can immediately begin using memory operations through the MCP protocol.

Key features

• Persistent memory storage via filesystem
• Read, write, and search operations for stored memories
• Background daemon for automatic memory consolidation
• Works with Claude Desktop, Cursor, Kiro, and any MCP client
• CLI-based setup and configuration
• Native consolidation and extraction of memory insights