For AI assistants

Plug Skamper into your AI

Skamper exposes its UK running-route database to AI assistants via the Model Context Protocol. Ask Claude or Cursor for "a flat 10k near Bath" or "a trail route in the Lake District" and it'll answer using real Skamper data, with links straight to the route page.

MCP server URL

Endpoint
https://skamper.com/mcp

Public, no API key required. Streamable HTTP transport.

What it can do

  • ✨ recommend_route — persona-shaped picks ("trail runner, 15k, avoid roads, scenic")
  • 🔍 search_routes — numeric filters: distance, ascent, surface %, course shape
  • 📄 get_route — full details, reviews and link to the route page
  • 🏃 find_5k — parkrun-distance routes near a location
  • 🏞️ list_national_parks / routes_in_national_park
  • 📚 browse_runlists / get_runlist — curated and user lists

Machine-readable manifest: /.well-known/mcp.json

Claude Desktop

Settings → Developer → Edit Config, then add Skamper to mcpServers and restart Claude.

claude_desktop_config.json
{
  "mcpServers": {
    "skamper": {
      "url": "https://skamper.com/mcp"
    }
  }
}

Cursor / Windsurf / Cline

Open MCP settings and add a new server with the URL above, or paste this config.

mcp.json
{
  "mcpServers": {
    "skamper": {
      "url": "https://skamper.com/mcp"
    }
  }
}

Try these prompts

  • "Find me a flat 10k near Cambridge using Skamper."
  • "Recommend a trail run under 15k in the Peak District."
  • "What's a good 5k near Bushy Park?"
  • "Show me Skamper's featured run lists."

A note on ChatGPT

ChatGPT (web/app) doesn't support external MCP servers yet. Claude Desktop, Cursor, Windsurf, Cline and Roo Code all do. We'll add ChatGPT the moment OpenAI ships support.