Your AI visibility data, inside your own tools
rankzupAI is more than a dashboard: with the MCP server and REST API you connect your scores, tracked queries and competitor data to Claude, Cursor, n8n or your own automation with a single key.
Which tools does the MCP server expose?
The MCP server exposes three read-only tools. Your assistant queries the data and reasons on top of it. Endpoint: https://api.rankzup.ai/mcp
list_projects
Lists the projects in your account: id, name, brand and site domain.
get_visibility
The latest visibility snapshot of a project: overall score, confidence band, per-engine scores, mention rate and share of voice.
list_prompts
Tracked queries: mentioned or not per engine, best position, from the latest scan.
Automate reporting with the REST API
Automate reporting through JSON endpoints with the same rz_ key: build your own dashboard, export to a warehouse or connect Looker Studio.
/api/v1/projects
Project list and single-project detail. Queries and the event feed live on separate endpoints.
/api/v1/projects/{id}/snapshots
Daily score snapshots: raw series for trend analysis and your own charts.
/api/v1/looker/snapshots
A flattened endpoint for the Looker Studio connector: client dashboards in minutes.
MCP and API access ships with the Scale and Agency plans. Generate your key from the panel in one click. Every endpoint is read-only, your data never changes.
What MCP means for your AI visibility data
MCP means your AI visibility data works in any assistant that supports the protocol, with no custom integration to build. The Model Context Protocol is an open-source standard for connecting AI applications to external systems. An MCP server can offer three kinds of building blocks: tools, resources and prompts. The rankzupAI server offers tools only, all of them read-only. Your assistant decides when to call them, based on what you ask.
MCP server or REST API
The MCP server suits questions you ask in an assistant, while the REST API suits code that runs on a schedule.
| MCP server | REST API | |
|---|---|---|
| Who makes the call | Your assistant, when your question needs data | Your code or a scheduled job |
| Best for | Ad hoc questions in Claude Code, Cursor or n8n | Dashboards, data warehouses, Looker Studio |
| What you write | A plain-language question | HTTP requests that return JSON |
| Key | rz_ key as a Bearer header | The same rz_ key |
Frequently asked questions
Do I need an API key to use the MCP server?
Yes. The MCP server authenticates with the same rz_ key as the REST API. You generate it in the panel in one click once you are on the Scale or Agency plan.
Do I need to write code to use it?
No. You paste one configuration block into your client, then ask questions in plain language. The assistant picks the right tool and passes the project id for you.
Which AI engines does the MCP data cover?
It covers the engines your project is measured on. Every plan includes Google AI Overviews, ChatGPT, Gemini and Perplexity, with Claude, Grok or Google AI Mode as optional add-ons. The get_visibility tool returns a separate score for each engine.
Get your key, connect your assistant
Start without a card and let the data accumulate, then get your key: move to Scale when you are ready and generate your rz_ key.
Published: Last updated: