Skills pipeworx-climate

Long-range climate projections from Open-Meteo — temperature trends from 1950 to 2050 with multi-model comparison

install
source · Clone the upstream repo
git clone https://github.com/openclaw/skills
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/openclaw/skills "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/brucegutman/pipeworx-climate" ~/.claude/skills/clawdbot-skills-pipeworx-climate && rm -rf "$T"
manifest: skills/brucegutman/pipeworx-climate/SKILL.md
source content

Climate Projections

What will the climate look like in 2040? This pack uses the Open-Meteo Climate API to retrieve long-term temperature projections for any location on Earth, spanning 1950 to 2050. Compare outputs across multiple climate models.

Tools

  • get_climate_projection
    — Temperature data for a location using the EC_Earth3P_HR model. Provide latitude, longitude, start date, and end date.
  • compare_models
    — Run the same query across multiple climate models simultaneously to see how projections diverge.

Use cases

  • Visualizing how average temperatures in a specific city have changed since 1950
  • Comparing climate model outputs for a research paper
  • Building climate risk assessments for real estate or agriculture
  • Answering "will summers in London be hotter in 2045?"

Example: temperature trend for Tokyo

curl -s -X POST https://gateway.pipeworx.io/climate/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_climate_projection","arguments":{"latitude":35.6762,"longitude":139.6503,"start_date":"2020-01-01","end_date":"2050-12-31"}}}'

Returns daily temperature projections with min, max, and mean values.

Setup

{
  "mcpServers": {
    "pipeworx-climate": {
      "command": "npx",
      "args": ["-y", "mcp-remote@latest", "https://gateway.pipeworx.io/climate/mcp"]
    }
  }
}

Tips

  • Date range must fall between 1950-01-01 and 2050-12-31
  • The
    compare_models
    tool is heavier — use smaller date ranges for faster responses
  • Latitude/longitude in decimal degrees (e.g., Tokyo: 35.6762, 139.6503)