M-Flow
A bio-inspired cognitive memory engine — a new paradigm for Graph RAG.
What is M-Flow?
M-Flow is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to bio-inspired cognitive memory engine — a new paradigm for graph rag.
A bio-inspired cognitive memory engine — a new paradigm for Graph RAG.
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- A bio-inspired cognitive memory engine — a new paradigm for
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx m-flowConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use M-Flow
M-Flow is a bio-inspired cognitive memory engine for AI agents that implements a new paradigm for Graph RAG (Retrieval-Augmented Generation) by modeling memory the way biological systems do — with episodic memory, associative recall, and a dynamic knowledge graph that evolves as agents interact. It provides an MCP server interface so AI assistants can store experiences, retrieve contextually relevant memories, build interconnected knowledge graphs, and reason over long-term accumulated information across sessions. Developers building agentic AI systems that need genuine long-term memory — not just a vector database — will find M-Flow a richer alternative to simple embedding stores.
Prerequisites
- Python 3.10+ installed
- An MCP-compatible client such as Claude Desktop or Cursor
- A graph database backend compatible with M-Flow (check the repository for supported backends)
- An embedding model API key if using semantic memory retrieval (e.g., OpenAI text-embedding-3-small)
Clone the M-Flow repository
Clone the repository from GitHub. Since there is no published PyPI package, installation is done from source.
git clone https://github.com/FlowElement-ai/m_flow.git
cd m_flowInstall dependencies
Install the Python dependencies using pip or uv. Review the requirements.txt or pyproject.toml in the repository for exact dependencies.
pip install -e .
# or with uv:
uv syncConfigure the memory engine
Set up your environment variables for the embedding provider and any graph database connection strings required by M-Flow. Check the repository's configuration documentation for the exact variable names.
export OPENAI_API_KEY=sk-your-key-here
export MFLOW_GRAPH_DB_URL=your-graph-db-connection-stringStart the MCP server
Launch M-Flow in MCP server mode so your AI client can connect to it and use the memory tools.
python -m m_flow.server
# or via npx wrapper:
npx m-flowConfigure your MCP client
Add M-Flow to your Claude Desktop or other MCP client configuration.
{
"mcpServers": {
"m-flow": {
"command": "npx",
"args": ["m-flow"]
}
}
}Store and retrieve your first memory
Once connected, ask your AI assistant to store an experience or fact using M-Flow. The memory engine will create nodes and edges in the knowledge graph representing the stored information.
M-Flow Examples
Client configuration
Claude Desktop configuration for the M-Flow memory engine MCP server.
{
"mcpServers": {
"m-flow": {
"command": "npx",
"args": ["m-flow"]
}
}
}Prompts to try
Sample prompts for using M-Flow's bio-inspired memory capabilities.
- "Remember that we decided to use PostgreSQL for the user service in the last architecture meeting"
- "What do you recall about our discussions on the payment integration?"
- "Show me the knowledge graph connections related to our microservices design decisions"
- "Store this insight: the rate limiter should be placed at the API gateway layer, not individual services"
- "What are all the things you remember about our project's authentication system?"Troubleshooting M-Flow
Repository not found or 404 when cloning
The repository at github.com/FlowElement-ai/m_flow may be private or have moved. Check the current status at the repository URL. If unavailable, use `npx m-flow` which may pull from an npm-published package.
Memory retrieval returns empty results despite stored memories
Semantic retrieval requires embeddings to be generated and indexed. Ensure your embedding API key (e.g., OPENAI_API_KEY) is set correctly and that the embedding service is reachable. Check the server logs for embedding generation errors.
Graph database connection errors on startup
Verify that the graph database is running and the connection URL in your environment variables is correct. M-Flow may support multiple backends — check the repository's README for the exact environment variable names and supported database configurations.
Frequently Asked Questions about M-Flow
What is M-Flow?
M-Flow is a Model Context Protocol (MCP) server that bio-inspired cognitive memory engine — a new paradigm for graph rag. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install M-Flow?
Follow the installation instructions on the M-Flow GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with M-Flow?
M-Flow works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is M-Flow free to use?
Yes, M-Flow is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
M-Flow Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to M-Flow? Here are other popular knowledge & memory servers you can use with Claude, Cursor, and VS Code.
MemPalace
★ 52.6kA local AI memory system that stores all conversations verbatim and organizes them into navigable structures. It provides 19 MCP tools for AI assistants to search and retrieve past decisions, debugging sessions, and architecture debates automatically
Kratos
★ 25.7k🏛️ Memory System for AI Coding Tools - Never explain your codebase again. MCP server with perfect project isolation, 95.8% context accuracy, and the Four Pillars Framework.
Context Mode
★ 15.4kAn MCP server that preserves LLM context by intercepting large data outputs and returning only concise summaries or relevant sections. It enables efficient sandboxed code execution, file processing, and documentation indexing across multiple programm
Memu
★ 13.7kMemory for 24/7 proactive agents like OpenClaw.
MemOS
★ 9.3kMemOS (Memory Operating System) is a memory management operating system designed for AI applications. Its goal is: to enable your AI system to have long-term memory like a human, not only remembering what users have said but also actively invoking, u
Everos
★ 5.4kBuild, evaluate, and integrate long-term memory for self-evolving agents.
Browse More Knowledge & Memory MCP Servers
Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up M-Flow in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use M-Flow?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.