> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fullreach.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# The MCP server

> Ask ChatGPT, Claude, Cursor or another AI assistant about this brand's figures, and get answers from the same data as the dashboard.

The FullReach AI MCP server lets an AI assistant read this organization's measurements: the same figures as the dashboard, and the answers behind them. Ask in plain language, and the assistant calls the tools that answer.

## What an assistant can do with it

* **Answer a question.** "Which competitors gained share of voice on Gemini this month?" The assistant picks the tools and the filters.
* **Run a ready-made analysis.** A [prompt](/mcp/prompts), such as a weekly pulse or an action plan, is a slash command in the assistant.
* **Work with your other tools.** The assistant can put the figures into a client report, a slide or a ticket. It uses the tools it already has.

[Use cases](/mcp/use-cases) lists questions to start from.

## How it works

1. The assistant connects to `https://mcp.fullreach.ai`. On connect, the server tells it what each figure means and which tool to call first.
2. The assistant calls [tools](/mcp/tools). Each tool reads the answers already collected. Most tools return the figures of one dashboard page, with the same filters.
3. The assistant can open the answers behind a figure, word for word, with `get_answer`.

Every tool reads and changes nothing. No tool asks a platform anything live, so a question costs no collection.

## Access

* **Who can connect.** Any organization with a project, on every tier. [Plan limits](/reference/limits) states the rate limit.
* **What a connection reads.** Every project of the organization, or one project. A key or a connection for one project reads nothing else.
* **How to connect.** Sign in and approve, or use an API key. [Connect a client](/mcp/connect) gives the steps for each assistant.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.