MCP Server Tool Calling with LangChain

A demonstration of using MCP (Model Control Protocol) servers with LangChain to create and manage custom tools for language models.

This project consists of three main components:

1. Client (client.py): The main application that uses LangChain and a Groq model to answer questions. It connects to the MCP servers to get a list of available tools.
2. Math Server (mathserver.py): An MCP server that provides basic arithmetic operations (add and multiply). It communicates over standard input/output.
3. Weather Server (weather.py): An MCP server that provides a simple weather lookup service. It communicates over HTTP.

Project Structure

• client.py: Main client application that connects to MCP servers and uses LangChain for tool calling.
• mathserver.py: MCP server providing math operations (add, multiply).
• weather.py: MCP server providing weather information.
• main.py: Simple entry point for the application.
• .env: For storing environment variables (e.g., GROQ_API_KEY).
• requirment.txt: Python dependencies.

Features

• Math Server: Provides two basic arithmetic tools:
  • add(a, b): Adds two numbers.
  • multiply(a, b): Multiplies two numbers.
• Weather Server: Provides a tool to get weather information:
  • get_weather(location): Returns a mock weather forecast for a given location.
• LangChain Integration: The client uses LangChain to create a ReAct agent that can intelligently use the tools provided by the MCP servers to answer questions.
• Multi-Protocol Communication: Demonstrates two different MCP transport protocols: stdio for the math server and streamable_http for the weather server.

Prerequisites

• Python 3.8+
• A Groq API key.

Setup and Running the Application

1. Clone the repository:

   git clone <repository-url>
   cd MCP-server-tool-calling-langChain
   

2. Install dependencies:

   pip install -r requirment.txt
   

3. Set up environment variables: Create a file named .env in the project root and add your Groq API key:

   GROQ_API_KEY="your-groq-api-key"
   

4. Run the Weather Server: In a terminal, run the weather server:

   python weather.py
   

   This will start an HTTP server on localhost:8000.

5. Run the Client: In a separate terminal, run the client application:

   python client.py
   

   The client will automatically start the mathserver.py as a subprocess and then connect to both the math and weather servers. It will then proceed to ask two questions to the LangChain agent to demonstrate the tool-calling capabilities. You should see output in the client's terminal showing the available tools, the response to the math question, and the response to the weather question.