Aviation data,
native to your AI.
Plug live SkyLink aviation data straight into Claude, Cursor, or any MCP-compatible client - every endpoint becomes a tool the model can call, with no glue code and no scraping.
What you get
Your AI, plugged into the sky
One config block connects your MCP client to SkyLink - then the model can pull weather, airports, ADS-B, charts, and more, just by asking.
Add one server block - works with any MCP-compatible client.
The model picks the right tool and calls it for you.
Every endpoint, a tool
The whole API, as callable tools. A sample of what the model can reach.
Start free - 1,000 requests/month.
Add your API key to the config and your AI is connected - works on the free tier.
Built for agents
A hosted bridge between your AI client and the full SkyLink v3.1 API - no servers to run, no integration code to write.
Works with any MCP client
Claude Desktop, Cursor, Cline, Windsurf, or anything that speaks the Model Context Protocol - paste one config block and you're connected.
Every endpoint as a tool
Weather, airports, ADS-B, charts, flight status, carbon, ML estimates - the whole v3.1 surface is exposed as tools the model can invoke.
Zero glue code
No SDK wiring, no request plumbing, no response parsing. The MCP layer handles tool schemas and calls - you just describe what you want.
Fully hosted
The server is a streamable HTTP endpoint we run - no self-hosting, no bridge process, no infrastructure to maintain. Point your client at the URL and go.
Natural-language queries
Ask in plain English - the model chooses the right endpoint, fills in the parameters, and grounds its answer in live SkyLink data.
Secure key handling
Your API key lives in the client's config header - sent as a transport header, never embedded in prompts or exposed to the model.
Setup
Add the server in one config
Paste one block into your MCP client's config, swap in your API key, and restart. The server is a hosted HTTP endpoint - nothing to run yourself.
Full Setup Guide{
"mcpServers": {
"skylink": {
"url": "https://data.skylinkapi.com/mcp",
"headers": { "x-api-key": "YOUR_API_KEY" }
}
}
}Replace YOUR_API_KEY · no self-hosting required
- 1Settings → Developer → Edit Config
- 2Paste into
mcpServers - 3Restart Claude Desktop
- 1Settings → MCP
- 2Add MCP Server
- 3Paste the config block
In practice
Then just ask
Once connected, your AI reaches for SkyLink tools on its own - picking the right endpoint, filling in parameters, and grounding its answer in live data.
No prompt engineering for API calls, no copy-pasting JSON. Just a question and a grounded answer.
MCP Documentation* Sample conversation for illustration only.
Connect once, then just ask - and get grounded answers from live aviation data.
Tools in reach
A sample of the v3.1 endpoints exposed as callable tools - the model picks what it needs.
Works everywhere
If it speaks the Model Context Protocol, it can connect to SkyLink.
Connection
One block to wire it up
Drop these into your client's mcpServers config and you're live.
Ready to connect your AI?
One config block, every SkyLink endpoint as a tool - for Claude, Cursor, or any MCP client. Add your API key and start asking.