Memory MCP Server

What is Memory MCP Server?

A Python implementation of an MCP server that processes and saves memory in the form of a knowledge graph. It enables AI systems to store, retrieve, and query structured data (entities, relations, observations) across sessions.

How to use Memory MCP Server?

Install via uv add mcp-server-memory or pip install mcp-server-memory. Run the server with the mcp-server-memory command. Optionally set the MEMORY_FILE_PATH environment variable to customize the memory file location. Configure for MCP clients like Claude or Cursor by adding the server definition to their settings files.

Key features of Memory MCP Server

• Create and manage entities with structured information
• Establish typed relations between entities
• Add observations to existing entities
• Delete entities, relations, and observations
• Search nodes in the knowledge graph
• Retrieve specific nodes by name

Use cases of Memory MCP Server

• Persist AI assistant memory across conversations
• Build and query a structured knowledge base from chat sessions
• Enable context recall by storing named entities and their relationships
• Facilitate long‑term agent memory in multi‑turn interactions

FAQ from Memory MCP Server

How do I install and run the server?

Install using uv add mcp-server-memory or pip install mcp-server-memory, then run mcp-server-memory. Optionally set the MEMORY_FILE_PATH environment variable for a custom storage location.

Where is the knowledge graph data stored?

By default, data is saved in a memory.json file in the package directory. You can change the path with the MEMORY_FILE_PATH environment variable.

What MCP tools does the server expose?

It exposes nine tools: create_entities, create_relations, add_observations, delete_entities, delete_observations, delete_relations, read_graph, search_nodes, and open_nodes.

What are the runtime dependencies?

The server requires Python and either uv or pip for installation. It uses the MCP (Model Control Protocol) interface.

Are there configuration examples for popular MCP clients?

Yes, the README provides example JSON configurations for both Claude (claude_desktop_config.json) and Cursor (mcp.json), using either uv or a pip‑installed Python module.