Document Search MCP Server

🚀 Python MCP Server - Document Search

This is a simple MCP (Minimal Command Protocol) server that allows users to search for documents from popular libraries such as LangChain, LlamaIndex, and OpenAI via the Serper API. The server uses HTTP requests and BeautifulSoup to retrieve search results from web pages and extract relevant document content.

🚀 Quick Start

This Python MCP server enables users to search for documents in popular libraries. To get started, follow the installation steps below and then you can use it to query the latest documentation for specific libraries.

✨ Features

• Support searches in the documentation of LangChain, LlamaIndex, and OpenAI.
• Use the Serper API to perform web searches.
• Extract and return the text content from relevant document pages.
• Can be used interactively via the MCP protocol.

📦 Installation

Step 1: Install Python Dependencies

To set up the project, clone this repository and install the required dependencies.

git clone <repository-url>
cd <repository-directory>
pip install -r requirements.txt

Or, if you don't use the requirements.txt file:

pip install httpx beautifulsoup4 mcp python-dotenv

Step 2: Set Environment Variables

Create a .env file in the project root directory and add your Serper API key to it.

SERPER_API_KEY=your-serper-api-key-here

Make sure to use the dotenv package to load the .env file.

Step 3: Run the Server

After setting up, run the server using the following command:

uv run main.py

The server will start and wait for input.

💻 Usage Examples

Basic Usage

Once the server is running, you can use it to query the latest documentation for specific libraries. Here is an example of a tool call:

Start the server using uv run main.py. In another terminal or from within an interactive MCP client, call the get_docs tool to search for documents:

get_docs("Chroma DB", "langchain")

This will search for "Chroma DB" in the LangChain documentation and return relevant content.

Available Libraries

• LangChain: Documentation available at langchain.com
• LlamaIndex: Documentation available at llamaindex.ai
• OpenAI: Documentation available at platform.openai.com

🔧 Technical Details

main.py

• FastMCP: Initializes the MCP server.
• Tools: The server has two tools:
  • dummy_tool: A simple tool to confirm that the server is running properly.
  • get_docs: Retrieves the latest documentation for a given query and library (LangChain, LlamaIndex, OpenAI).
• search_web(): Handles searches using the Serper API.
• fetch_url(): Retrieves a URL and uses BeautifulSoup to extract the text content.

dotenv and logging

• Loads the Serper API key from the .env file.
• Provides debug-level logging for better understanding of server operations and potential issues.