Kanxueshan MCP
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Kanxueshan MCP这个周末去子梅垭口,能看到贡嘎山日照金山吗?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
看雪山 MCP · Kanxueshan MCP
🏔️ 日照金山(雪山金光)预报的 MCP 服务器 —— 在 Claude、Cursor、Cherry Studio、Claude Desktop 等任何 MCP 客户端里,直接问:
「这个周末去子梅垭口,能看到贡嘎山日照金山吗?」 「明天在飞来寺,能看到卡瓦格博吗?」 「30.9, 102.8 附近有哪些雪山可以看?」
返回:日出/日落金光时段、0-100 可见度评分、云量/湿度/风速因子,以及观景点建议。

直接用(推荐,无需安装)
托管端点(streamable HTTP,免费、无需 API Key):
https://www.yilong.art/mcpClaude Code / 命令行
claude mcp add --transport http kanxueshan https://www.yilong.art/mcpCursor(~/.cursor/mcp.json)
{
"mcpServers": {
"kanxueshan": { "url": "https://www.yilong.art/mcp" }
}
}其他客户端:在「MCP / 连接器」设置里添加远程 HTTP 服务器,地址填 https://www.yilong.art/mcp。
图文接入指引:https://www.yilong.art/ai.html ・ 网页版:https://www.yilong.art
Related MCP server: China Weather MCP Server
工具(Tools)
工具 | 说明 | 参数 |
| 搜索中国雪山及观景点(子梅垭口、飞来寺、猫鼻梁…) |
|
| 按经纬度查周边可见雪山 + 当日日出/日落金山预报 |
|
| 某山峰未来 1-7 天预报 |
|
数据与算法
气象:Open-Meteo 云量(含低云)、降水概率、湿度、风速/阵风
地形:DEM 高程 + 视线分析(判断某观景点能否看到该峰)
金光时段:太阳高度角(Golden Hour)+ 当地时区
评分:综合云量/低云/湿度/风,0-100
覆盖中国主要雪山与观景点(贡嘎山、四姑娘山、卡瓦格博、玉珠峰、慕士塔格……),并提供多语种山峰名检索。
自托管(Self-host)
本仓库是 MCP 接入层(约 100 行),需要自行准备一份看雪山 API:
pip install -r requirements.txt
KXS_API=http://127.0.0.1:3000 python3 mcp_server.py # 默认监听 0.0.0.0:3999环境变量:KXS_API(后端地址)· HOST / PORT(监听,生产建议 HOST=127.0.0.1 走反代)·
KXS_TOOL_LOG(可选,设为文件路径则每次工具调用追加一行 JSONL:工具名/参数/来源 IP/耗时/成败,便于自建用量统计)·
KXS_STATS_DB(可选,设为 SQLite 路径则每次调用在 toolcall 表写一行,字段同上,便于直接查询/做面板)。
两个都不设则完全不写盘。
English
Kanxueshan MCP — golden-hour sunrise/sunset visibility forecasts for snow mountains in China.
Ask your MCP client: "Can I see the golden light on Gongga Mountain from Zimei Pass this weekend?" It returns the golden-hour windows (sunrise/sunset), a 0-100 visibility score, weather factors (cloud cover, humidity, wind), and suggested viewpoints.
Hosted endpoint (no install, no API key):
https://www.yilong.art/mcpTools: search_peak(name) · query_peaks(lat, lng, date?) · forecast_peak(peak, days?, viewpoint?, date?)
Data: Open-Meteo weather + a terrain/viewshed engine (DEM, solar altitude, line-of-sight). Forecast responses are in Chinese; peak names accept both Chinese and English queries.
Web app: https://www.yilong.art ・ Setup guide: https://www.yilong.art/ai.html
License
TBD
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for current weather and multi-day forecasts worldwide, Chinese city names and output.
Smarter Weather MCP: forecasts, alerts, outlooks, observations, AQI, grids, and map imagery.
Sunrise-Sunset MCP — wraps the sunrisesunset.io API (free, no auth)
Hosted MCP server for Xweather weather data: conditions, forecasts, alerts, and more.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceProvides comprehensive weather data including forecasts, air quality, weather alerts, historical data, and astronomical information through the Hefeng Weather API. Supports real-time weather, multi-day forecasts, hourly predictions, and specialized data like sunrise/sunset times and precipitation forecasts.4MIT
- FlicenseNot gradedqualityDmaintenanceEnables querying weather forecasts (1-7 days) and meteorological warnings for Chinese cities using the QWeather API. Supports detailed weather data including temperature, humidity, wind, precipitation, UV index, and real-time weather alerts.2-
- AlicenseNot gradedqualityDmaintenanceAn MCP server that provides real-time weather information, 4-day forecasts, and city search functionality for Chinese cities via the AMap API. It enables users to query weather data using city names or administrative codes through natural language interactions.1MIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server that analyzes upcoming weather at alpine peaks and trailheads worldwide, ranking the best windows for climbing and ascent using high-resolution forecast models.36 npm3MIT