Probonobonobo Sui MCP Server

🚀 MCP Server

The MCP (Message Communication Protocol) server is a fast API service based on FastAPI and FAISS, primarily designed to support RAG (Retrieval Augmented Generation) applications. This document details how to build and use the server.

🚀 Quick Start

📦 Installation

Run the following command in the project's root directory to install all necessary dependencies:

pip install -r requirements.txt

⚙️ Configuration

Copy the .env.example file and rename it to .env, then modify the configuration as needed:

GITHUB_TOKEN=your_github_token_here
OPENAI_API_KEY=your_openai_api_key_here

✨ Features

🔍 Directory Structure

mcp_server/
├── __init__.py             # Package initialization file
├── main.py                # Main server file
├── mcp_api.py             # MCP API implementation file
├── index_move_files.py    # Document indexing tool
├── local_query.py         # Local query tool
├── download_move_files.py # GitHub Move file extractor
├── rag_integration.py     # LLM integration module
├── pyproject.toml         # Package configuration file
├── requirements.txt       # Dependency management file
├── .env.example           # Environment variable example file
├── README.md              # Project documentation
├── data/                  # Data storage directory
├── docs/                  # Example document directory
│   └── move_files/        # Directory for downloaded Move files
├── models/                # Model implementation directory
│   └── vector_store.py    # FAISS vector store implementation
└── utils/
    ├── document_processor.py  # Document processing tool
    └── github_extractor.py    # GitHub file extraction tool

💻 Core Functional Modules

1. MCP API Implementation (mcp_api.py)

Provides the following main interfaces:

• /mcp/action: Supports two operation types, retrieve_documents and index_documents.

Request Example

curl -X POST "http://localhost:8000/mcp/action" -H "Content-Type: application/json" -d '{"action_type": "retrieve_documents", "payload": {"query": "What is RAG?", "top_k": 3}}'

2. Document Indexing (index_move_files.py)

Converts documents into text and stores them in the FAISS index:

python index_move_files.py --docs-dir docs/move_files --index-file data/faiss_index.bin

3. Local Query (local_query.py)

Retrieves relevant documents directly from the index:

python local_query.py "What is RAG?"

📋 Use Cases

Case: RAG Pipeline Implementation

The complete RAG workflow is as follows:

1. Search Query: The user submits a question.
2. Retrieval: The system finds relevant documents from the vector database.
3. Context Building: Formats the retrieved documents into a prompt.
4. LLM Generation: Sends the prompt to the LLM model, combined with the retrieved content.
5. Enhanced Response: The LLM generates an answer based on the provided information.

This workflow is fully implemented in rag_integration.py.

🚀 Project Expansion

Feature Expansion Suggestions

• Support more data sources (e.g., S3, HDFS, etc.).
• Implement distributed indexing.
• Add a document pre - processing pipeline.
• Integrate multiple LLM providers.
• Add a result caching mechanism.
• Support incremental index updates.

How to Contribute

1. Fork this repository.
2. Create a feature branch.
3. Submit changes.
4. Open a Pull Request.

📚 Documentation

GitHub Repository

📄 License

LICENSE

📞 Contact

• Email: contact@example.com
• GitHub: https://github.com/yourusername