Skip to main content

MCP Overview

What is MCP?

Model Context Protocol (MCP) is an open protocol for connecting AI applications with external tools, data sources, and services.

Through MCP, an MCP client can discover and invoke tools provided by an MCP server. RAGFlow can either act as an MCP server to expose its capabilities to external clients or connect to external MCP servers to extend the tools available to RAGFlow Agents.

MCP usage scenarios in RAGFlow

RAGFlow currently supports the following two MCP usage scenarios.

Use RAGFlow as an MCP server

RAGFlow can act as an MCP server and expose capabilities such as knowledge base retrieval to external MCP clients.

The RAGFlow MCP server runs as a standalone component alongside a properly functioning RAGFlow server.

The MCP server supports two launch modes:

  • Self-host mode: Provide a RAGFlow API key when starting the MCP server. The MCP server uses this API key to access RAGFlow resources available to the corresponding user.
  • Host mode: The MCP server is not bound to a fixed user. Each MCP client provides its own authentication information in requests and can access only the RAGFlow resources available to that user.

The RAGFlow MCP server supports the following transports:

  • Streamable HTTP: Communicates through the /mcp endpoint.
  • SSE: Communicates through the /sse and /messages/ endpoints and is retained for compatibility with legacy MCP clients.

A typical request flow is:

External MCP Client
        ↓
RAGFlow MCP Server
        ↓
RAGFlow Server
        ↓
RAGFlow capabilities

For information about starting and configuring the RAGFlow MCP server, see Launch RAGFlow MCP Server.

For details about the MCP tools exposed by RAGFlow and their parameters, see RAGFlow MCP Tools.

For examples of calling the RAGFlow MCP server with Python or curl, see RAGFlow MCP Client Examples.

Connect an external MCP server to RAGFlow

RAGFlow can also connect to third-party or self-hosted MCP servers and discover the MCP tools they provide.

Once configured, these tools can be used by RAGFlow Agents, allowing agents to interact with external systems and services.

A typical request flow is:

External MCP Server
        ↓
      RAGFlow
        ↓
       Agent

For information about configuring, testing, and using external MCP servers, see Connect an External MCP Server to RAGFlow.

Choose the right MCP scenario

Choose an MCP usage scenario based on what y