Ragflow Knowledge MCP Server

🚀 RAGFlow Knowledge Base MCP Server Configuration Guide

RAGFlow is a question - answering system based on large language models, aiming to help users retrieve knowledge and answer questions through natural language. To ensure the normal operation of RAGFlow, it is necessary to correctly configure the MCP (Message Control Plane) server. This document will provide a detailed introduction on how to configure the RAGFlow Knowledge Base MCP server.

🚀 Quick Start

This guide will lead you through the steps of installing dependencies, configuring the MCP server, using the RAGFlow knowledge base, advanced configuration, and troubleshooting.

✨ Features

• Knowledge Retrieval: Empowers users to conduct knowledge retrieval and question - answering via natural language.
• Flexible Configuration: Allows for custom configuration of the MCP server according to actual requirements.
• Knowledge Base Management: Enables creation, import, and management of knowledge bases.

📦 Installation

1.1. Install Python

Ensure that the latest version of Python 3 is installed:

python --version

If installation is required, visit the Python official website.

1.2. Install pip

Ensure that pip is installed and updated to the latest version:

pip install --upgrade pip

1.3. Install Dependencies

Run the following command to install the required dependency packages for the RAGFlow Knowledge Base MCP server:

pip install -r requirements.txt

The requirements.txt file should contain all necessary Python packages.

💻 Usage Examples

2.1. Download the Configuration File

Download the MCP server configuration file, usually named config.yaml, from the RAGFlow official website.

2.2. Modify the Configuration File

Modify the config.yaml file according to actual needs. Here is an example configuration:

server:
  name: "RAGFlow Knowledge Base"
  transport: "stdio"  # Or choose "sse"
  sse_port: 41106
  timeout: 60
  log:
    console_enabled: true
    file_enabled: true
    log_level: "INFO"
    log_file: "~/logs/mcp.log"

2.3. Start the MCP Server

Run the following command to start the MCP server:

python mcp_server.py --config config.yaml

If using the stdio mode, you can interact with the server through standard input.

3.1. Create a Knowledge Base

Use the tools provided by RAGFlow to create or import a knowledge base:

ragflow create - knowledge - base --name my_knowledge_base

3.2. Import Data

Import documents or data into the knowledge base:

ragflow import --knowledge - base my_knowledge_base --documents /path/to/documents

3.3. Conduct Q&A

Perform Q&A through the MCP server:

curl -X POST http://localhost:41106/ask \
  -H "Content - Type: application/json" \
  -d '{"question": "有什么问题？"}'

🔧 Technical Details

4.1. Log Configuration

Configure log output through environment variables or the config.yaml file:

export SIMP_LOGGER_LOG_LEVEL="DEBUG"
export SIMP_LOGGER_LOG_FILE="/path/to/mcp.log"

4.2. Extended Configuration

If you need to extend functionality, refer to RAGFlow's plugin mechanism or contact the official support team.

📚 Documentation

5.1. Check Logs

View the log file for more information:

tail -f ~/logs/mcp.log

5.2. Update Dependencies

Ensure that all dependencies are updated to the latest version:

pip install --upgrade .

5.3. Contact Support

If the problem remains unresolved, contact the RAGFlow support team.

📄 License

No license information provided in the original document, so this section is skipped.

Conclusion

By following the above steps, you can successfully configure and run the RAGFlow Knowledge Base MCP server. If you have any questions or need further assistance, refer to the official documentation or contact technical support.