MCP server

Connect an AI assistant to Supermetry analytics through its MCP endpoint.

Verify the MCP endpoint before adding it to an MCP client.

Create an API key with stats:read and access to only the sites the assistant should query.

Set the endpoint and key.

export SUPERMETRY_URL="https://analytics.example.com"
export SUPERMETRY_API_KEY="replace-with-your-key"

Initialize a JSON-RPC session.

curl --fail-with-body -X POST "$SUPERMETRY_URL/api/public/v1/mcp" \
  -H "Authorization: Bearer $SUPERMETRY_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"curl","version":"1"}}}'

List the available tools.

curl --fail-with-body -X POST "$SUPERMETRY_URL/api/public/v1/mcp" \
  -H "Authorization: Bearer $SUPERMETRY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "MCP-Protocol-Version: 2025-06-18" \
  --data '{"jsonrpc":"2.0","id":2,"method":"tools/list"}'

The server exposes list_sites, get_summary, get_timeseries, get_breakdown, get_realtime, list_goals and run_funnel.

What you should see: initialization names the server supermetry, and tools/list returns the seven analytics tools.

MCP server

Related