peak-window-mcp-app
Alpine weather window analyzer — MCP App powered by Open-Meteo
Install / Use
claude mcp add ByteOverDev -- npx -y github:ByteOverDev/peak-window-mcp-appIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
OtherSupported Platforms
Skill content
View source on GitHubPeakWindow MCP App
An MCP App that analyzes upcoming weather at alpine peaks and trailheads worldwide, ranking the best windows for climbing and ascent. Uses the best available forecast model for each region: GeoSphere Austria AROME (2.5 km) for Central Europe, MeteoSwiss ICON-CH2 (2 km) for the Alpine region, Météo-France AROME (2.5 km) for France, and Open-Meteo globally (~11 km).
<p align="center"> <a href="https://byteoverdev.github.io/peak-window-mcp-app/"> <img src="screenshot.png" alt="PeakWindow UI" width="700" /> </a> </p>What it does
- Fetches hourly forecasts (temperature, precipitation, wind, gusts, cloud cover, snow line) from the best available provider for the requested coordinates
- Automatically selects the highest-resolution model: GeoSphere (2.5 km, Central Europe), MeteoSwiss (2 km, Alpine region), Météo-France (2.5 km, France), or Open-Meteo (~11 km, global)
- Falls through to the next provider on API failure for resilience
- Lapse-corrects temperatures to summit elevation using model terrain derived from surface pressure (-6.5 °C/km)
- Scores each hour against alpine climbing thresholds and identifies contiguous good-weather windows
- Serves an interactive UI (React, uPlot charts) as an MCP App resource with horizon tape, mountain profile, and Ventusky-style chart panels
Requirements
This is an MCP App — it requires an MCP host that supports the Apps protocol to render its interactive UI. Compatible hosts include:
- Claude Desktop (macOS / Windows)
- ChatGPT (via OpenAI Apps SDK)
- Any chat client implementing the MCP Apps spec
The server runs as a standard MCP server over stdio or SSE. The rich UI (charts, horizon tape, mountain profile) renders inline in any compliant host. Without one you still get the scored text output.
Live UI demo — fetches real forecasts in the browser; pick any peak to see its scored windows and real DEM silhouette.
Setup
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"peak-window": {
"command": "npx",
"args": ["-y", "peak-window", "--stdio"]
}
}
}
Then ask Claude something like: "What's the best weather window to climb Großglockner this week?"
HTTP server
npx peak-window # starts on http://localhost:3001/mcp
Development
git clone https://github.com/ByteOverDev/peak-window-mcp-app.git
cd peak-window-mcp-app
npm install
npm run dev # hot reload (UI + server)
npm run serve:stdio # stdio transport
MCP tool
peak-window — provide lat, lon, and optionally peakName and summitElevationM. Returns scored hours, top weather windows, and a full time-series payload rendered by the embedded UI.
Data sources
| Provider | Resolution | Coverage | Via | |---|---|---|---| | GeoSphere Austria | 2.5 km | Central Europe (5.5–22.1°E, 43–51.8°N) | Direct API | | MeteoSwiss ICON-CH2 | 2 km | Alpine region (0.5–16.5°E, 43–49.9°N) | Open-Meteo | | Météo-France AROME | 2.5 km | France & surrounds (-9–14°E, 38–55°N) | Open-Meteo | | Open-Meteo | ~11 km | Global | Direct API |
Related Skills
Agent-Reach
84.2kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
73.4kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
ruflo
73.0k🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
CowAgent
47.1kOpen-source super AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
