FastMCP Server with LangGraph Agent & JWT Auth

A project to explore FastMCP, LangChain's LangGraph v1.0, and JWT authentication for MCP. This project implements a simple todo application using Model Context Protocol (MCP) server with an AI-powered interface.

Features

• Task Management API with FastMCP
• AI Agent Interface using LangChain and Groq
• JWT Authentication
• Simple In-Memory Task Storage
• Interactive CLI Interface

Prerequisites

• uv package package manager
• Python 3.13 or higher
• FastMCP 2.12.5 or higher
• LangChain 1.0.0 or higher
• Groq API access

Installation

1. Clone the repository

2. Install uv (if not already installed):

   curl -LsSf https://astral.sh/uv/install.sh | sh
   

3. Create a virtual environment and install dependencies:

   uv sync
   

Configuration

Create a .env file in the root directory with the following variables:

GROQ_API_KEY=your_groq_api_key
JWT_PUBLIC_KEY=your_jwt_public_key
JWT_ALGORITHM=HS256  # or your preferred algorithm
JWT_TOKEN=your_jwt_token

Project Structure

• mcp_server.py - Main FastMCP server implementation
• tools.py - Task management tools and temporary database
• agent.py - AI agent implementation using LangChain and Groq
• test_client.py - Example client implementation

Usage

1. Start the MCP Server:

   uv run mcp_server.py
   

2. Run the AI Agent Interface:

   uv run agent.py
   

3. For testing, you can use the test client:

    uv run test_client.py
   

Available Tools

• add_task: Add a new task to the system
• get_tasks: Retrieve all tasks
• remove_task: Remove a specific task by ID

Dependencies

• fastmcp >= 2.12.5
• langchain >= 1.0.0
• langchain-groq >= 1.0.0
• langchain-mcp-adapters >= 0.11.1
• langgraph >= 1.0.0
• python-dotenv >= 1.1.1

Development

The project uses Python 3.13 and follows standard Python project structure. Make sure to:

1. Update dependencies in pyproject.toml as needed