Weather MCP

A FastMCP server that provides weather data from Open-Meteo and renders an interactive weather dashboard through MCP Apps.

Features

• Search locations by city or place name

• Get current conditions and up to 16 forecast days

• Switch between Celsius and Fahrenheit

• Display weather descriptions and condition symbols

• Cache location and forecast responses for five minutes

• Render the dashboard directly inside compatible MCP hosts

Related MCP server: bridgemcp-server-weather

Requirements

• Python 3.14+

• uv

• An MCP host with MCP Apps support for the interactive dashboard

Setup

Install the project dependencies:

uv sync

No weather API key is required for normal non-commercial Open-Meteo usage.

Run The Server

Run the server over the default stdio transport:

uv run python main.py

Inspect the registered tools:

uv run fastmcp list main.py

Preview The Dashboard

FastMCP includes a local browser preview for app tools:

uv run fastmcp dev apps main.py

Open the displayed local URL and select weather_dashboard. The dashboard supports searching for another location and switching temperature units.

MCP Tools

search_locations

Find matching locations and return their coordinates, country, region, and time zone.

get_weather

Get current weather and a daily forecast for a location.

Arguments:

• location: City or place name

• forecast_days: Number of forecast days from 1 to 16

• temperature_unit: celsius or fahrenheit

weather_dashboard

Open the interactive weather dashboard with the same weather arguments.

Project Structure

main.py                         MCP server entry point
src/tools.py                    MCP tool registration and response shaping
src/services/weather_service.py Open-Meteo client and response cache
src/formatting.py               UI-friendly forecast formatting
src/dashboard.py                Prefab MCP App UI

Data Source

Weather data is provided by Open-Meteo, which combines forecasts from multiple national weather services. Review the Open-Meteo terms and licensing before commercial deployment.