Mcp Server Weather Example

Deploying & Connecting Your Remote MCP Weather Server

This project lets you deploy a cloud-based Model Context Protocol (MCP) server that delivers real-time weather data using the U.S. National Weather Service API. It’s built with FastAPI, supports Server-Sent Events (SSE) for real-time communication, and uses secure API key authentication.

This repository provides a ready-to-deploy Weather MCP server. You'll walk through the full setup from cloning the code, configuring authentication, deploying to Azure Container Apps, and finally connecting it inside Visual Studio Code using the MCP extension.

---

Clone the Repository

git clone https://github.com/your-username/mcp-weather-server.git
cd mcp-weather-server

---

Understand the Structure

Before deploying, familiarize yourself with the core components:

• weather.py – Defines two tools: get_alerts() and get_forecast() using the U.S. National Weather Service API.
• main.py – Wraps the tools in a FastAPI app exposed at /sse, with support for SSE.
• api_key_auth.py – Implements x-api-key header-based security for incoming requests.
• Dockerfile – Builds and packages the app using Python 3.13 and the uv web server.

---

Configure API Key Authentication

Set your API key(s) securely via an environment variable:

export API_KEYS=<YOUR_API_KEY>

You can also supply multiple comma-separated keys. No need to hardcode secrets into the source code.

---

Build & Deploy to Azure Container Apps

Use the Azure CLI to build and deploy in a single step:

az containerapp up \
  -g <RESOURCE_GROUP> \
  -n weather-mcp \
  --environment <CONTAINERAPPS_ENV_NAME> \
  -l westus \
  --env-vars API_KEYS=<YOUR_API_KEY> \
  --source .

This command will:

• Build the image using the included Dockerfile.
• Deploy it to Azure Container Apps.
• Inject your API key(s) via --env-vars.

Once deployed, Azure CLI will return a public URL this is the endpoint your MCP clients will use.

---

Connect from Visual Studio Code

With the server up, connect it inside VS Code:

1. Create a .vscode/mcp.json file in your project:

{
  "inputs": [
    {
      "type": "promptString",
      "id": "weather-api-key",
      "description": "Weather API Key",
      "password": true
    }
  ],
  "servers": {
    "weather-sse": {
      "type": "sse",
      "url": "https://weather-mcp.orangebush.westus.azurecontainerapps.io/sse",
      "headers": {
        "x-api-key": "${input:weather-api-key}"
      }
    }
  }
}

2. Open the MCP extension sidebar in VS Code.
3. Click Start next to your server entry.
4. Enter the API key when prompted.

You're now connected and ready to process real-time weather queries.

---

Try It with GitHub Copilot Chat (Agent Mode)

1. Launch GitHub Copilot Chat in Agent mode.
2. Ask something like:

"How's the weather forecast in Georgia?"

The agent will call your get_forecast tool behind the scenes and return a live forecast based on NWS data.

---

You're All Set!

Your MCP server is now:

• Cloud-hosted
• API-key protected
• Integrated with VS Code and GitHub Copilot

You can now build custom agents that interact with real-time weather data from anywhere.