Follow these steps to run the OpenWeather MCP agent and toolkit at the Room level for testing in the cloud.

Environment variables required:

• OWM_API_KEY: OpenWeatherMap API Key
  (Obtain from https://openweathermap.org/api)

​

Steps

1. Install Meshagent with all dependencies:

   Copy

   Ask AI

   pip install "meshagent[all]"
   

2. Sign up and authenticate with Meshagent:

   Follow instructions at: https://docs.meshagent.com/cli/getting_started

3. Start the OpenWeather MCP server as an agent in a Meshagent Room:

   Copy

   Ask AI

   meshagent service test \
     --room=test \
     --role=agent \
     --image=meshagent/mcp-openweather:latest \
     --env MESHAGENT_PORT=8001 \
     --env OWM_API_KEY=<OWM_API_KEY> \
     --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
     --name=mcp-openweather-service-test
   

   • This command will start a Meshagent Room named test with the OpenWeather MCP server available. Rooms close after inactivity.

4. Start a chatbot in the same Room with access to the OpenWeather toolkit:

   Copy

   Ask AI

   meshagent chatbot join \
     --room=test \
     --agent-name=mcp-openweather \
     --name=mcp-openweather \
     --toolkit=mcp-openweather
   

   • This command adds a chatbot agent with the OpenWeather toolkit to your Room.
   • The CLI will output a link to the Room.

5. Test it out:

   Open the Room link in your browser to interact with the agent and try out the OpenWeather MCP tools.

​

Project Level Deployment

Establish persistent MCP services and chatbot so they are automatically present in every Room in your Meshagent project—ideal for production scenarios.

1. Deploy the OpenWeather MCP service as an agent at the project level:

   Copy

   Ask AI

   meshagent service create \
     --role=agent \
     --image=meshagent/mcp-openweather:latest \
     --env MESHAGENT_PORT=8001 \
     --env OWM_API_KEY=<OWM_API_KEY> \
     --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
     --name=mcp-openweather-service
   

2. Deploy the Meshagent chatbot integrated with the OpenWeather toolkit:

   Copy

   Ask AI

   meshagent service create \
     --image="meshagent/cli:latest" \
     --port="num=9001 path=/agent liveness=/ type=meshagent.callable participant_name=mcp-openweather-chatbot" \
     --env="MESHAGENT_PORT=9001" \
     --name="mcp-openweather-chatbot-service" \
     --command="meshagent chatbot service --agent-name=mcp-openweather-chatbot --toolkit=mcp-openweather"
   

No need to manage separate MCP server/chatbot processes locally; these services will join any new project-created Room automatically.

​

Tools Available

​

Openweather MCP Server

A simple MCP service that provides current weather and 5-day forecast using the free OpenWeatherMap API.

Tools provided by this Server:

​

Tools Details

​

Tool: weather

Get current and forecast weather information for a specific City.

This tool may perform destructive updates.

This tool interacts with external entities.

​

Meshagent Resources

• Meshagent: https://www.meshagent.com/
• Documentation: https://docs.meshagent.com
• OpenWeather MCP Docker Image: https://hub.docker.com/r/meshagent/mcp-openweather