Skip to main content
AI Growth Agent helps companies get recommended in AI search — Google AI Overviews, Google organic results, and ChatGPT — through autonomous programmatic content generation. AI Growth Agent MCP is a headless Model Context Protocol (MCP) server that lets AI clients — Claude Cowork, Claude Code, Codex, Cursor — work with your AI Growth Agent accounts, reporting, content planner, and Google Search Console performance through natural language. Ask a question in your client and it calls the right tools for you. See Concepts for the domain model behind the answers — the search universe, citations, and brand mentions. Endpoint: https://mcp.aigrowthagent.co/mcp

Authentication

The server is an OAuth 2.1 protected resource. Every user connects with their own bearer token: on first use your client opens a browser sign-in on the AI Growth Agent login page, and every tool call then acts as you.
  • Your credentials are typed only into the browser — never through the AI.
  • Access tokens are short-lived (1 hour) and refreshed automatically by the client. To sign in again, reconnect or re-add the server.
  • There is no login tool; authentication happens at connection time.

Set up your client

Try it

Once connected, ask your client:
  • “How is my account performing against competitors?”
  • “Which domains are most cited across my search universe?”
  • “Where am I winning in AI Overviews this week?”
  • “What share of my organic search traffic comes from AI Growth Agent content?” (Google Search Console — requires adding [email protected] to your GSC property with Full access; see Google Search Console (GSC))
Citation and mention answers cover your tracked universe — the campaign and standard seed terms — and the most-cited domains and most-mentioned brands come back as a top 10 plus your own domain or brand pinned for reference. Read reading counts and ranks correctly before quoting one of those numbers. See the task guides for listing accounts and finding AI Overview wins, the concepts page, the tool reference, common workflows, and the guides — server prompts and resources, the first-analysis runbook, and the GSC data-fidelity and analysis-guardrails rule sets.