Use SnowSure inside Claude, ChatGPT, Cursor & more
One URL. No install. Real-time snow conditions and 14-day forecasts for 300+ resorts, available in any AI chat that supports the Model Context Protocol.
MCP server URL
https://www.snowsure.ai/mcpStreamable HTTP · stateless · public · read-only
300+ resorts
From the Alps to Niseko, Whistler, Aspen, the Andes, and beyond.
7 weather models
ECMWF, GFS, GEM, JMA, ICON, Météo-France, Met Norway — averaged into one forecast.
11 tools, 4 resources
Powder rankings, multi-model forecasts, regional summaries, trip planning.
Add to Claude.ai
Available on Claude.ai (web) on the Pro, Max, Team, and Enterprise plans. Custom connectors are currently in beta.
- Open claude.ai and go to Settings → Connectors.
- Click Add custom connector.
- Name it
SnowSureand paste the URL:https://www.snowsure.ai/mcp. - Leave OAuth Client ID and Secret blank — SnowSure is public and read-only.
- Click Add.

Claude’s “Add custom connector” dialog. Leave the OAuth fields blank.
Then: in any chat, look for the tools / connectors menu near the composer (often a slider or wrench icon) and toggle SnowSure on for that conversation. You only need to do this once per chat.
Add to ChatGPT
Available in the ChatGPT desktop app (and recent web builds) for Plus, Pro, Team, and Enterprise plans. Look for “Connectors”, “Custom GPT connectors”, or “MCP servers” depending on the build.
- Open ChatGPT → Settings → Connectors (or MCP servers).
- Click Add HTTP server / Add custom connector.
- Name it
SnowSure, pastehttps://www.snowsure.ai/mcp, and save. - Authentication: None.
If you don’t see Connectors in your build yet, you can also use SnowSure through our existing Custom GPT.
Add to Cursor
Cursor supports both HTTP and stdio MCP servers. We recommend HTTP (no install).
- Open Cursor → Settings → Cursor Settings → MCP & Integrations.
- Click Add new MCP server.
- Choose URL / HTTP, name it
snowsure, pastehttps://www.snowsure.ai/mcp. - Save and toggle the server on.
Or paste this directly into your .cursor/mcp.json:
{
"mcpServers": {
"snowsure": {
"url": "https://www.snowsure.ai/mcp"
}
}
}Try it: MCP Inspector & curl
Want to poke at the server before wiring it into a client? Two quick options:
MCP Inspector (visual)
npx @modelcontextprotocol/inspector # then paste https://www.snowsure.ai/mcp into the URL field
curl (raw JSON-RPC)
curl -X POST https://www.snowsure.ai/mcp \
-H 'Content-Type: application/json' \
-H 'Accept: application/json, text/event-stream' \
-d '{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "find_best_powder",
"arguments": { "minSnowfall": 10, "limit": 5 }
}
}'Legacy: stdio package
For clients that don’t yet support remote MCP (older Claude Desktop builds, some self-hosted agents), the same tools are available as a local npm package.
{
"mcpServers": {
"snowsure": {
"command": "npx",
"args": ["-y", "snowsure-mcp-server"]
}
}
}Source on GitHub. MIT-licensed.
What can I ask?
Once connected, just chat normally. The AI will pick the right tool automatically. Some ideas to get started:
Discovery
- “Where’s the best snow right now?”
- “Top 5 ski resorts in Europe by SnowSure score.”
- “Which resorts in Japan got the most fresh snow in the last 24 hours?”
- “Find me powder in the Alps with at least 15cm overnight.”
Specific resorts
- “What are conditions at Niseko Hanazono?”
- “Tell me about Jackson Hole — depth, recent snow, forecast.”
- “How’s Val d’Isère looking for next week?”
- “Show me the live webcams for Cervinia.”
Forecast confidence
- “Compare the 14-day forecasts for Whistler and Aspen.”
- “How reliable is the forecast for Zermatt? Show me model agreement.”
- “Which model is most bullish on snow for St. Anton this week?”
Trip planning
- “I’m an advanced skier looking at the Alps in late February. Where should I go?”
- “Find resorts with at least 200cm base, 2,500m+ summit, and SnowSure score above 50.”
- “Plan a powder week in Japan for an intermediate skier.”
Tip: if the AI doesn’t reach for the connector, prefix your question with “using SnowSure, …” and it’ll call a tool instead of guessing from training data.
Tools & resources exposed
The connector exposes 11 tools that the AI can call, plus 4 resources for raw data access.
Tools
get_snow_reportTop-ranked resorts by SnowSure score, recent snowfall, depth, or forecast.
get_resortComprehensive resort details: conditions, multi-model forecast, history, webcams.
search_resortsFind resorts by name, country, or region.
find_best_powderResorts with the freshest powder right now (24h snowfall).
compare_forecastsSide-by-side comparison of all 7 weather models with confidence rating.
get_weather_forecastDay-by-day weather for up to 14 days for any resort.
find_resorts_by_criteriaFilter by depth, elevation, runs, score, country, region.
get_snow_historySeason totals plus 5-year and 30-year averages.
plan_ski_tripAI-powered trip recommendations for a region, ability level, and dates.
get_webcam_statusLive webcam links and thumbnails.
get_regional_summaryRegional / country statistics and top resorts.
Resources
snowsure://snow-reportCurrent global snow report (top 50 resorts).
snowsure://resortsAll 300+ resorts with current conditions.
snowsure://regionsRegion/country breakdown with resort counts.
snowsure://api-docsOpenAPI specification summary.
Troubleshooting
›Claude says “Couldn’t reach the MCP server”
Make sure the URL has no trailing slash or extra characters. Our endpoint is exactly https://www.snowsure.ai/mcp. If the error persists, try removing the connector and re-adding it — Claude caches the failure state.
›The AI ignores my SnowSure question and uses training data
In Claude.ai, make sure SnowSure is toggled on in the per-chat tools menu. Prefixing your prompt with “using SnowSure…” also works.
›Resort search returns nothing
Resort names sometimes differ from common spelling (e.g. “matterhorn-ski-paradise” for Zermatt). Ask the AI to search first — it'll return the correct slug to use for follow-ups.
›Is there an authentication option?
Not currently. The connector is intentionally public and read-only, mirroring our public REST API at /api/v1. If you need authenticated / write access for an integration, email api@snowsure.ai.
›What about rate limits?
No hard limits today; please be reasonable. We log usage and may add per-IP throttling if abuse becomes an issue. For high-volume use cases, talk to us.
Building something with SnowSure?
We’d love to hear about it. Whether you’re an AI agent author, a ski-trip planner, or a resort partner — tell us how the connector is showing up.