Azure Container Apps remote MCP server example

This MCP server uses SSE transport and is authenticated with an API key.

---

Project Overview

This project is a FastAPI-based Model Context Protocol (MCP) server designed for Azure Container Apps. It provides weather-related tools (such as alerts and forecasts) via a secure, API-key-protected Server-Sent Events (SSE) interface. The server is intended for integration with MCP-compatible clients, such as Visual Studio Code or custom scripts.

Setup Instructions

1. Clone the repository:

   git clone https://github.com/KRVaghani/Azure-MCP-Server.git
   cd Azure-MCP-Server
   

2. Install prerequisites:
   • Python 3.11 or later
   • uv
3. Set up the virtual environment and dependencies:

   uv venv
   uv sync
   

4. Set the API key:

   export API_KEYS=devkey  # or your chosen key
   

5. Run the server:

   uv run fastapi dev main.py
   

Client Usage

A sample Python client (model_request_client.py) is included for command-line interaction:

python model_request_client.py

• Enter your API key (default: devkey)
• Choose a tool: get_alerts (by US state) or get_forecast (by latitude/longitude)
• Enter the required parameters
• The server response will be displayed in your terminal

API Endpoints

• /sse (GET): Main SSE endpoint for MCP protocol communication. Requires x-api-key header.
• /messages (SSE): Used internally for message streaming (session-based).

Example curl Test

curl -H "x-api-key: devkey" http://localhost:8000/sse

Troubleshooting

• If you see ModuleNotFoundError, ensure your virtual environment is activated and dependencies are installed.
• If the client hangs, check that the server is running and the API key matches.
• For pip issues in the venv, run:

  python -m ensurepip --upgrade
  python -m pip install --upgrade pip
  

Contributing

Pull requests are welcome! Please open an issue first to discuss major changes.

License

MIT License (add LICENSE file if needed)

Running locally

Prerequisites:

• Python 3.11 or later
• uv

Run the server locally:

uv venv
uv sync

# linux/macOS
export API_KEYS=<AN_API_KEY>
# windows
set API_KEYS=<AN_API_KEY>

uv run fastapi dev main.py

VS Code MCP configuration (mcp.json):

{
    "inputs": [
        {
            "type": "promptString",
            "id": "weather-api-key",
            "description": "Weather API Key",
            "password": true
        }
    ],
    "servers": {
        "weather-sse": {
            "type": "sse",
            "url": "http://localhost:8000/sse",
            "headers": {
                "x-api-key": "${input:weather-api-key}"
            }
        }
    }
}

Deploy to Azure Container Apps

az containerapp up -g <RESOURCE_GROUP_NAME> -n weather-mcp --environment mcp -l westus --env-vars API_KEYS=<AN_API_KEY> --source .

If the deployment is successful, the Azure CLI returns the URL of the app. You can use this URL to connect to the server from Visual Studio Code.

If the deployment fails, try again after updating the CLI and the Azure Container Apps extension:

az upgrade
az extension add -n containerapp --upgrade