MCP Memory Server - Python Implementation

MCP Memory Server - Python Implementation is an MCP server that provides persistent knowledge graph storage and retrieval using standard JSON Lines (JSONL) files. Built as a Python port of the official Model Context Protocol TypeScript memory server, it connects AI assistants to a local graph database of entities, relationships, and observations. Developers and AI practitioners use this server to equip language models with long-term memory across sessions without relying on complex external database engines. The server exposes nine core tools covering graph manipulation, such as creating entities, establishing relations, attaching observations, deleting elements, searching nodes by query text, and traversing interconnected nodes. Because it uses identical JSONL structures, users can migrate existing memory files directly from the TypeScript reference implementation. The server communicates over standard input and output (stdio) and supports custom file paths configured through environment variables.

Category: AI Memory & Context

Tags: jsonl, knowledge-graph, persistence, retrieval

Visit MCP Memory Server - Python Implementation

How to install and configure MCP Memory Server - Python Implementation

1. Clone the repository and navigate into the folder: bash git clone https://github.com/jason-c-dev/memory-mcp-server-py cd memory-mcp-server-py 2. Create and activate a Python virtual environment: bash python3 -m venv .venv source .venv/bin/activate 3. Install dependencies: bash pip install -r requirements.txt 4. Add the server to your Claude Desktop configuration file (claude_desktop_config.json) under mcpServers: json { "mcpServers": { "memory": { "command": "python", "args": ["/path/to/mcp_memory_server.py"], "env": { "MEMORY_FILE_PATH": "/path/to/memory.json" } } } } Replace /path/to/ with the absolute path to your script and desired storage location.

What you can do with MCP Memory Server - Python Implementation

• Storing persistent user identity, preferences, and communication habits across separate conversational sessions in Claude Desktop. * Mapping complex professional networks by creating entities for organizations and linking colleagues via typed relationship edges. * Recording sequential project observations and technical milestones associated with specific code repositories or tools. * Searching through stored knowledge graph nodes using free-text queries to recall context during technical troubleshooting. * Inspecting connected subgraphs via the open_nodes tool to analyze direct relationships between people, projects, and organizations.

Key facts

• Open Source
• https://github.com/jason-c-dev/memory-mcp-server-py
• AI Memory & Context, Notes & Knowledge Management
• jsonl, knowledge-graph, persistence, retrieval

Part of MCP Servers

How do I install MCP Memory Server - Python Implementation?

Clone the GitHub repository, set up a Python 3.8 or higher virtual environment, and run pip install -r requirements.txt. Once dependencies are installed, you can configure your MCP client such as Claude Desktop or Cursor to launch mcp_memory_server.py using the virtual environment Python interpreter.

What can MCP Memory Server - Python Implementation do?

It provides nine tools to build and query a local knowledge graph. The server lets clients create, delete, and read entities, establish relations between nodes, record observations, search entity properties by text query, and explore connected nodes using a local JSONL file.

Which MCP clients work with MCP Memory Server - Python Implementation?

The server works with any MCP client supporting stdio communication. Tested setups include Claude Desktop, Cursor IDE, and AWS Q CLI. Each client requires specifying the python executable path, the server script path, and an optional environment variable for the memory file.

What operating systems are supported?

This implementation was developed and tested exclusively on macOS. While it operates on standard Python and should function on other Unix-like systems, it has not been officially tested on Windows or Linux environments.

Is MCP Memory Server - Python Implementation compatible with the TypeScript version?

Yes, it provides full feature parity with the official TypeScript implementation. It implements the exact same nine MCP tools and utilizes identical JSON Lines file formats, allowing you to reuse existing memory.json files without manual conversion.