Integration · MCP (AI assistants)

MCP (AI assistants) and Tellane.

Ask your assistant which lane has the worst drop-off, or to draft follow-ups for this week's high-intent leads, and it reads the answer from Tellane.

Plan: Every plan.

Exactly what moves

What the integration does

  • Tools: workspace_info, list_lanes, get_lane (stops and screens), analytics_summary, list_leads.
  • Transport: Streamable HTTP, stateless JSON, at POST {API}/mcp, authenticated with an API key in the Authorization header.
  • Read-only: a test key (tl_test_…) is enough; no tool edits anything.
  • Rate-limited like any key: 600 requests a minute with X-RateLimit-* headers.

Setup

Connect it

  1. Settings → API keys: create a test key.
  2. Settings → Integrations → MCP: copy the config block for your assistant (Claude Desktop, Claude Code, Cursor, ChatGPT, Copilot share the same shape).
  3. Paste it into the assistant's MCP configuration and ask it for your workspace info to confirm.

Reference

MCP config (shape)

{
  "mcpServers": {
    "tellane": {
      "url": "https://api.tellane.com/mcp",
      "headers": { "Authorization": "Bearer tl_test_…" }
    }
  }
}

Full detail in the API reference.

FAQ

Questions

Can the assistant change a lane?

No. Every tool is read-only by design.

Which assistants work?

Anything that speaks MCP over Streamable HTTP with custom headers — Claude Desktop, Claude Code, Cursor, ChatGPT, GitHub Copilot.

More integrations

Also connects to