Sunshine Atlas MCP server

A remote, read-only Model Context Protocol server that answers "where is it sunny?" with data: monthly 0–100 Sunshine Scores, day/night temperatures, rainfall and sea temperature for 3,833 destinations — every one with its own airport — from long-term climate normals. It serves the same open dataset published at sunshineatlas.com/data (CC BY 4.0, DOI 10.5281/zenodo.21322408).

Live endpoint (no key, no signup):

Code

https://sunshineatlas.com/api/mcp

Listed in the official MCP registry as com.sunshineatlas/sunshine-atlas. Docs page: sunshineatlas.com/mcp.

Tools

Tool	Answers
`find_sunny_destinations`	"Where is it sunny (and warm) in November?" — ranked by that month's Sunshine Score, filterable by continent/country, minimum day temperature, swimmable sea (≥21 °C), population
`get_destination_climate`	"When is the best time to visit Faro?" — full 12-month climate card for one destination (city, "city, country" or IATA code)
`compare_destinations`	"Algarve or Crete in October?" — side-by-side comparison of 2–5 destinations with a one-line verdict

Every result carries attribution and citable sunshineatlas.com destination URLs.

Connect a client

Claude Code

Terminal

claude mcp add --transport http sunshine-atlas https://sunshineatlas.com/api/mcp

Claude (custom connector), ChatGPT connectors, or any Streamable-HTTP client — add the URL https://sunshineatlas.com/api/mcp.

Generic JSON config (Cursor, etc.)

config.json

{
  "mcpServers": {
    "sunshine-atlas": {
      "url": "https://sunshineatlas.com/api/mcp"
    }
  }
}

Self-host

The server is a single stateless Node file (fresh McpServer per request, plain JSON responses, binds to 127.0.0.1 — put it behind your reverse proxy). It reads one dataset file and hot-reloads it on mtime change.

Terminal

npm install
curl -O https://sunshineatlas.com/data/sunshine-atlas-destinations.json
PORT=8787 DATA_FILE=./sunshine-atlas-destinations.json node server.mjs

GET /healthz reports status and destination count. server.json is the MCP-registry manifest.

Data & methodology

Sunshine Score = 100 × warmth × (0.5 + 0.5 × (0.55 × dryness + 0.45 × sunniness)), computed from long-term climate normals — methodology.
Sources: temperature, rainfall and sunshine from NASA POWER climatology (2001–2020); places & populations from GeoNames & OurAirports; sea temperatures via Open-Meteo Marine.
Dataset: sunshineatlas.com/data · CSV/JSON · CC BY 4.0 · mirrors on GitHub, Hugging Face, Kaggle.

License

Code MIT. Data CC BY 4.0 (credit "Sunshine Atlas" + link). Not a forecast — climate normals describe typical conditions, not this week's weather.