Weather MCP
Provides 12 weather tools for conditions, forecasts, alerts, model imagery, radar, and 205 meteorological calculations, enabling weather data access and analysis through Claude Code.
README
Weather MCP
MCP server with 12 weather tools — conditions, forecasts, alerts, NWP model imagery, NEXRAD radar, and 205 meteorological calculations. Built for Claude Code.
Data tools work immediately with just requests. Image rendering and calculations use optional Rust-backed packages for professional-grade output.
Quick Start
git clone https://github.com/FahrenheitResearch/weather-mcp
cd weather-mcp
pip install requests mcp
Add to your Claude Code MCP config (~/.claude/settings.json):
{
"mcpServers": {
"weather": {
"command": "python",
"args": ["/full/path/to/weather-mcp/server.py"]
}
}
}
Restart Claude Code. You now have 7 weather data tools available immediately.
Tools
Data (Python — always available)
| Tool | What it does |
|---|---|
wx_conditions |
Current obs: temperature, wind, sky, dewpoint |
wx_forecast |
NWS 7-day or hourly forecast |
wx_alerts |
Active warnings/watches with GeoJSON polygons for map overlays |
wx_metar |
Raw/decoded METAR for any ICAO station |
wx_brief |
Conditions + forecast + alerts in one call |
wx_global |
Non-US locations via Open-Meteo |
wx_severe |
SPC Day 1 outlook + active watches |
Images (Rust — optional)
| Tool | What it does |
|---|---|
wx_model_image |
NWP field rendered as PNG — 22+ products, 3 models, batch support |
wx_radar_image |
NEXRAD Level 2 reflectivity (1024px, 200km, dark background) |
wx_storm_image |
Radar with storm cell analysis |
Calculations (Rust — optional)
| Tool | What it does |
|---|---|
wx_calc |
Any of 205 meteorological functions (dewpoint, CAPE, LCL, Fosberg FWI, etc.) |
wx_sounding |
Model sounding at a point — 40 levels + all derived parameters |
Optional: Image & Calculation Packages
# Model imagery + soundings
pip install metrust cfrust rusbie rustweather
# Radar (build from source)
cd /path/to/rustdar
cargo build --release --bin radar-render
# Set path if not at ~/rustdar/
export RADAR_RENDER_PATH=/path/to/radar-render
What Can I Build With This?
Just tell Claude Code what you want. Examples:
- "Build me a fire weather dashboard with Red Flag Warning overlays" — Claude Code will use
wx_alerts(event_type="Red Flag Warning")and overlay the GeoJSON polygons on your map - "Show me the HRRR CAPE forecast for Oklahoma" —
wx_model_image(var="cape", lat=35.5, lon=-97.5, radius_km=300) - "Calculate the Fosberg Fire Weather Index for 95°F, 8% RH, 25 mph wind" —
wx_calc(function="fosberg_fire_weather_index", ...) - "What's the severe weather outlook?" —
wx_severe() - "Build a weather app with current conditions and radar" — Claude Code chains
wx_conditions+wx_radar_image
Credits
- Color tables: Solarpower07
- Meteorological calculations: metrust — 205 functions verified against MetPy
- GRIB decoding: cfrust (pure Rust, replaces eccodes)
- NWP downloads: rusbie (byte-range .idx filtering)
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。