mcp-remote-demo

This MCP server hosts three demo tool groups: locker, weather, and show_photo. Each group exposes a set of tools that you can invoke remotely through the MCP framework. The server expects authentication via an Authorization header using a Bearer token, and you can pass a profile ID in requests when needed. For testing and usage, you can connect to the server using the Streamable HTTP transport and call the registered tools (e.g., get_available_lockers_from_zone, get_alerts, get_forecast, show_photo) by sending the appropriate tool name, parameters, and profile context. The weather demo restricts lookups to US locations and fetches data from api.weather.gov; the locker and show_photo demos return hardcoded or demonstrative results to illustrate tool responses. If you’re building a client profile, you can configure multiple MCP tool servers in a profile array and route requests to the mcp-remote-demo server as needed.

To use a tool, construct a request that specifies the tool name and its parameters. For example, get_available_lockers_from_zone requires a zone parameter; get_alerts and get_forecast fetch weather-related data from the weather API for a given location, and show_photo demonstrates returning an image URL. When integrating into a client, ensure the Authorization header is included with the Bearer token and supply the profile ID where required by the server context.

Prerequisites:\n- Python 3.8+ installed on your system\n- Git installed to clone the repository\n\n1) Clone the repository:\n git clone https://github.com/your-org/mcp-remote-demo.git\n cd mcp-remote-demo\n\n2) (Optional) Create and activate a virtual environment:\n python3 -m venv venv\n source venv/bin/activate # On Windows use: venv\Scripts\activate.bat\n\n3) Install dependencies:\n pip install -r requirements.txt # If a requirements file is provided\n # If there is a setup.py or pyproject.toml, you can also run: pip install -e .\n\n4) Run the MCP server:\n python -m mcp_remote_demo_server # Ensure the module name matches your package\n\n5) Verify the server is running:\n curl -H "Authorization: Bearer <your_token>" http://localhost:8080/mcp/ # or the configured port/endpoint\n\nNotes:\n- Adjust the module name in the run command if your packaging uses a different entry point.\n- If you deploy with a different port or host, update your client configuration accordingly.\n\n

Tips and common considerations:\n- The three demos (locker, weather, show_photo) are designed for demonstration purposes with simple, hardcoded or mocked outputs. When integrating into production, you can replace or extend tool implementations with real data sources.\n- Authentication is sent in the Authorization header as a Bearer token; you can use the AuthTokenAuthProvider approach shown in the repository examples.\n- The weather tools are US-location specific; ensure you pass valid location queries to get relevant results.\n- If you run into issues connecting to the MCP server, verify that the Python entry point module name matches the installed package and that the server process has the necessary network permissions.\n- The transport option supported is streamable-http as indicated by the README, but you can adapt your client to any compatible MCP transport if needed.