Travel Itinerary Backend System with MCP

Travel Itinerary Backend System with MCP

A FastAPI-based backend system for managing travel itineraries. This project implements the Model Context Protocol (MCP) for enhanced AI assistant integration.

Features

• Database Architecture: SQLAlchemy models for itineraries, accommodations, transfers, and activities
• RESTful API: Endpoints to create and view trip itineraries
• Seed Data: Pre-populated database with realistic data for Phuket and Krabi regions
• Recommended Itineraries: Sample itineraries ranging from 2-8 nights
• MCP Integration: Full Model Context Protocol support for AI assistants

Installation

1. Ensure you have Python 3.8+ installed

2. Install uv (fast Python package installer) from here:

3. Clone this repository and navigate to the project directory:

   git clone https://github.com/aryanranderiya/itinerary-mcp-server
   cd itinerary-mcp-server
   

4. Create a virtual environment and install dependencies:

   uv venv
   uv sync
   

5. Activate the virtual environment:

   On Linux/macOS:

   source .venv/bin/activate
   

   On Windows (Command Prompt):

   .venv\Scripts\activate.bat
   

   On Windows (PowerShell):

   .venv\Scripts\Activate.ps1
   

Running the Project

Development Mode

Run the server in development mode with auto-reload: Only run in development mode if you need to edit the code.

fastapi dev

Production Mode

Run the server in production mode:

fastapi start

The API will be available at http://localhost:8000 and the API documentation at http://localhost:8000/docs.

Using the Model Context Protocol (MCP)

This project implements the Model Context Protocol (MCP), which enables AI assistants to interact with your API directly. This means AI tools can understand your API's capabilities, data structures, and execute operations on your behalf.

MCP Configuration

The project is already configured with the necessary MCP setup:

• fastapi_mcp Library: We use fastapi_mcp library to integrate MCP functionality with FastAPI
• mcp.json: The configuration is located in the .vscode/mcp.json file, which defines the MCP server settings:

  {
    "servers": {
      "stay-often-mcp1": {
        "type": "sse",
        "url": "http://localhost:8000/mcp"
      }
    }
  }
  

  This configuration connects to the MCP endpoint at http://localhost:8000/mcp using Server-Sent Events (SSE).

No additional setup is required to use the MCP functionality - it's built into the server and ready to use.

Supported MCP Clients

You can use any MCP-compatible client to interact with this API. MCP clients include:

• GitHub Copilot
• Claude Desktop
• And other MCP-compatible assistants

For more information about MCP, visit the official MCP documentation.

Example: Using with GitHub Copilot

Here's how you can interact with this API using GitHub Copilot:

1. Start the server:

   fastapi dev
   

2. In your conversation with GitHub Copilot, you can request operations like:

   • Get all itineraries:

     Can you show me all the recommended itineraries for Phuket?
     

   • Get a specific itinerary:

     Show me the details for itinerary ID 3
     

   • Create a new itinerary:

     Create a new 5-night itinerary in Krabi with the following details: Name: "Krabi Adventure", day 1 at hotel 5 with activities 10 and 11, day 2 at hotel 5 with activity 12, etc.
     

GitHub Copilot will understand these requests and execute the appropriate API calls on your behalf through the MCP protocol.

Screenshots