# Weather MCP Server This is a sample MCP Server in Python implementing weather tools with mock responses. It can be used as a scaffold for your own MCP Server. It includes the following features: - **Weather Tool**: A tool that provides mocked weather information based on the given location. - **Connect to Agent Builder**: A feature that allows you to connect the MCP server to the Agent Builder for testing and debugging. - **Debug in [MCP Inspector](https://github.com/modelcontextprotocol/inspector)**: A feature that allows you to debug the MCP Server using the MCP Inspector. ## Get started with the Weather MCP Server template > **Prerequisites** > > To run the MCP Server in your local dev machine, you will need: > > - [Python](https://www.python.org/) > - (*Optional - if you prefer uv*) [uv](https://github.com/astral-sh/uv) > - [Python Debugger Extension](https://marketplace.visualstudio.com/items?itemName=ms-python.debugpy) ## Prepare environment There are two approaches to set up the environment for this project. You can choose either one based on your preference. > Note: Reload VSCode or terminal to ensure the virtual environment python is used after creating the virtual environment. | Approach | Steps | | -------- | ----- | | Using `uv` | 1. Create virtual environment: `uv venv`
2. Run VSCode Command "***Python: Select Interpreter***" and select the python from created virtual environment
3. Install dependencies (include dev dependencies): `uv pip install -r pyproject.toml --extra dev` | | Using `pip` | 1. Create virtual environment: `python -m venv .venv`
2. Run VSCode Command "***Python: Select Interpreter***" and select the python from created virtual environment
3. Install dependencies (include dev dependencies): `pip install -e .[dev]` | After setting up the environment, you can run the server in your local dev machine via Agent Builder as the MCP Client to get started: 1. Open VS Code Debug panel. Select `Debug in Agent Builder` or press `F5` to start debugging the MCP server. 2. Use Microsoft Foundry Toolkit Agent Builder to test the server with [this prompt](vscode://ms-windows-ai-studio.windows-ai-studio/open_prompt_builder?model_id=github/gpt-4.1-mini&system_prompt=You%20are%20a%20weather%20forecast%20professional%20that%20can%20tell%20weather%20information%20based%20on%20given%20location&user_prompt=What%20is%20the%20weather%20in%20Shanghai?&track_from=vsc_md&mcp=weather_mcp). Server will be auto-connected to the Agent Builder. 3. Click `Run` to test the server with the prompt. **Congratulations**! You have successfully run the Weather MCP Server in your local dev machine via Agent Builder as the MCP Client. ![DebugMCP](https://raw.githubusercontent.com/microsoft/windows-ai-studio-templates/refs/heads/dev/mcpServers/mcp_debug.gif) ## What's included in the template | Folder / File| Contents | | ------------ | -------------------------------------------- | | `.vscode` | VSCode files for debugging | | `.aitk` | Configurations for Microsoft Foundry Toolkit | | `src` | The source code for the weather mcp server | ## How to debug the Weather MCP Server > Notes: > - [MCP Inspector](https://github.com/modelcontextprotocol/inspector) is a visual developer tool for testing and debugging MCP servers. > - All debugging modes support breakpoints, so you can add breakpoints to the tool implementation code. | Debug Mode | Description | Steps to debug | | ---------- | ----------- | --------------- | | Agent Builder | Debug the MCP server in the Agent Builder via Microsoft Foundry Toolkit. | 1. Open VS Code Debug panel. Select `Debug in Agent Builder` and press `F5` to start debugging the MCP server.
2. Use Microsoft Foundry Toolkit Agent Builder to test the server with [this prompt](vscode://ms-windows-ai-studio.windows-ai-studio/open_prompt_builder?model_id=github/gpt-4.1-mini&system_prompt=You%20are%20a%20weather%20forecast%20professional%20that%20can%20tell%20weather%20information%20based%20on%20given%20location&user_prompt=What%20is%20the%20weather%20in%20Shanghai?&track_from=vsc_md&mcp=weather_mcp). Server will be auto-connected to the Agent Builder.
3. Click `Run` to test the server with the prompt. | | MCP Inspector | Debug the MCP server using the MCP Inspector. | 1. Install [Node.js](https://nodejs.org/)
2. Set up Inspector: `cd inspector` && `npm install`
3. Open VS Code Debug panel. Select `Debug SSE in Inspector (Edge)` or `Debug SSE in Inspector (Chrome)`. Press F5 to start debugging.
4. When MCP Inspector launches in the browser, click the `Connect` button to connect this MCP server.
5. Then you can `List Tools`, select a tool, input parameters, and `Run Tool` to debug your server code.
| ## Default Ports and customizations | Debug Mode | Ports | Definitions | Customizations | Note | | ---------- | ----- | ------------ | -------------- |-------------- | | Agent Builder | 3001 | [tasks.json](.vscode/tasks.json) | Edit [launch.json](.vscode/launch.json), [tasks.json](.vscode/tasks.json), [\_\_init\_\_.py](src/__init__.py), [mcp.json](.aitk/mcp.json) to change above ports. | N/A | | MCP Inspector | 3001 (Server); 5173 and 3000 (Inspector) | [tasks.json](.vscode/tasks.json) | Edit [launch.json](.vscode/launch.json), [tasks.json](.vscode/tasks.json), [\_\_init\_\_.py](src/__init__.py), [mcp.json](.aitk/mcp.json) to change above ports.| N/A | ## Feedback If you have any feedback or suggestions for this template, please open an issue on the [Microsoft Foundry Toolkit GitHub repository](https://github.com/microsoft/vscode-ai-toolkit/issues)