MemFlow MCP

MemFlow MCP is an MCP server that connects Model Context Protocol clients like Claude Desktop to an external Memory Bank API for persistent context storage. Designed for developers and AI assistant users who require long-term context retention across multiple chat sessions, it allows large language models to store notes, retrieve historical data, and run semantic queries over saved information. The server communicates via standard MCP interfaces to manage memories with optional metadata tags and filtering. By hooking into an external backend supporting memory creation and retrieval endpoints, MemFlow MCP ensures that context is preserved outside of temporary chat windows. Users can recall prior instructions, maintain dynamic knowledge bases, and let AI agents perform semantic matching over recorded experiences. It requires an active Memory Bank API instance and Node.js to execute, functioning as a bridge between model reasoning and structured memory persistence.

Category: AI Memory & Context

Tags: embeddings, memory, persistence, retrieval, vector-search

Visit MemFlow MCP

How to install and configure MemFlow MCP

1. Ensure Node.js 18 or higher is installed and your Memory Bank API server is running at your target address (e.g., http://localhost:3000). 2. Open your Claude Desktop configuration file. 3. Add the server to the mcpServers block using your absolute path to npx (find it using which npx): json { "mcpServers": { "memflow": { "command": "/your/npx/path", "args": ["-y", "memflow-mcp"], "env": { "MEMBANK_API_URL": "http://localhost:3000", "MEMBANK_API_KEY": "your-api-key" } } } } 4. Alternatively, install globally using pnpm add -g memflow-mcp. 5. Restart Claude Desktop to complete setup.

What you can do with MemFlow MCP

• Saving user preferences, recurring constraints, and ongoing task notes directly from Claude conversations using the addMemory tool with custom tags. - Performing semantic search across previously stored memories to retrieve relevant conversation context when answering specific user prompts. - Browsing historical interactions and cataloged knowledge bases by listing stored memories with structural filtering options. - Maintaining continuity across multiple distinct chat sessions by querying external memory records using natural language requests.

Key facts

• https://github.com/sacmii/memflow-mcp
• AI Memory & Context, Notes & Knowledge Management
• embeddings, memory, persistence, retrieval, vector-search

Part of MCP Servers

How do I install MemFlow MCP?

You can run MemFlow MCP directly via npx by specifying your npx binary path in your Claude Desktop configuration file. Alternatively, install it globally using pnpm add -g memflow-mcp. You must also supply the MEMBANK_API_URL environment variable pointing to your active Memory Bank API server.

What can MemFlow MCP do?

MemFlow MCP enables AI models to retain persistent information across different sessions. It exposes three primary tools: addMemory to store content with optional tags, searchMemory to run semantic matching across saved items, and listMemories to browse stored records with filters.

What are the requirements to run MemFlow MCP?

The server requires Node.js version 18 or higher and a running Memory Bank API instance that supports standard POST and GET endpoints for memory handling. An API key can optionally be passed through the MEMBANK_API_KEY environment variable.

Is MemFlow MCP open source?

Yes, MemFlow MCP is open source software released under the MIT license. You can inspect the source code, modify it, or contribute to its development on its GitHub repository.