Meteosource has released a new Model Context Protocol (MCP) server that allows AI assistants to access comprehensive weather data without complex OAuth flows. The server, available at mcp.meteosource.com, speaks the Streamable HTTP transport and requires just two configuration values: the server address and a Meteosource API key. This low-friction setup is designed to get agents online quickly, with the specific API plan determining which tools are exposed to the model.

Simple Integration for Major AI Clients

The setup process is deliberately minimal. Developers can connect the server to Claude Code, Claude Desktop, Codex, and other clients by adding the endpoint https://mcp.meteosource.com/mcp and passing the API key via an Authorization: Bearer header or an X-API-Key field. Meteosource emphasizes that the key is never stored in full on their end and is used solely to call their API on behalf of the user. For teams concerned with security, the README provides specific instructions on keeping keys out of version control.

Seven Tools Spanning Weather and Energy

The server exposes seven distinct tools, though availability depends on the user's subscription tier. Standard weather plans grant access to search_places, get_forecast (including minute-by-minute precipitation), get_alerts for the USA, EU, and Canada, get_weather_history, and get_air_quality. The get_forecast tool supports up to 20 locations per call, which is a useful optimization for batch processing. For energy-focused agents, the Renewables plan unlocks get_solar_production and get_wind_production, offering output data for PV systems and wind farms from 20 years back to four days ahead.

Key Takeaways

  • The server uses Streamable HTTP transport and requires no OAuth sign-in, simplifying agent authentication.
  • Tool availability is tiered: standard weather tools are on Free/Startup/Standard plans, while renewable energy data requires the Renewables plan.
  • Integration guides for Claude Code, Codex, and team environments are hosted on GitHub.
  • The API key is transmitted via headers and is not stored in full by the server, with caching handled per key.

The Bottom Line

Meteosource's approach strips away the typical integration friction for weather data, making it a viable plug-in for any agent stack that needs real-world environmental context. The inclusion of solar and wind production data specifically targets the growing niche of energy-aware AI applications, distinguishing it from generic weather APIs.