> For the complete documentation index, see [llms.txt](https://docs.adpage.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.adpage.io/data-and-insights/marketing-integrations/ai-assistants-mcp.md).

# AI assistants (MCP)

Use your AdPage data inside Claude, ChatGPT or Manus and ask questions in plain language.

Your AdPage data, available inside the AI assistant you already use. Ask a question in plain language and it queries your containers, tracking data and connected platforms directly.

These are the only integrations that work at **account level** rather than per container — connect once and every container you have access to comes with it.

### What it can do

Once connected, the assistant can search your containers, query your server-side tracking, run data quality checks, and pull reporting from the platforms you have connected. In practice that means asking things like "which campaigns lost conversions last week and why" and getting an answer built from your own data rather than a guess.

With [Google Search Console](https://docs.adpage.io/data-and-insights/marketing-integrations/google-search-console.md) connected, it answers SEO questions too: which search queries rose or fell, which keywords are close to the top 3, which pages compete for the same keyword, and whether a page is indexed by Google — and if not, why.

Access follows your AdPage account. The assistant can only reach what you can reach, and it reads — anything that would change a setting goes through an explicit approval step.

### Claude

Add the AdPage MCP server to Claude Desktop or Claude Code:

```
https://mcp.adpage.io/mcp
```

The first time you use it, Claude asks you to sign in to AdPage through OAuth. There is **no token to copy or paste** — if a set of instructions has you pasting a key into a config file, it is out of date.

The **Claude** tile on the Integrations page has the exact configuration for your setup. A full walkthrough is in the FAQ: [Connect Claude to the Data Assistant](https://docs.adpage.io/faq/marketing-optimization/connect-claude-to-the-data-assistant).

<figure><img src="https://1954565367-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FK5kjEu8mwstNkOrvFl6S%2Fuploads%2FfvjM7YD3Ilpt0MsnUkSt%2Fintegrations-claude-tile-en.webp?alt=media" alt="Claude (MCP) tile opened with setup instructions for Claude Desktop"><figcaption><p>Your personal MCP settings on the Claude tile.</p></figcaption></figure>

### ChatGPT

AdPage is available as an app in ChatGPT. Open it from the **ChatGPT** tile on the Integrations page, which links straight to the AdPage app, and sign in with your AdPage account when prompted.

### Manus

Manus connects through the same MCP server. Because Manus runs tasks autonomously, access is arranged per account: use the request button on the **Manus AI** tile and support will set it up.

<figure><img src="https://1954565367-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FK5kjEu8mwstNkOrvFl6S%2Fuploads%2Fvy1RXLlc3Lctzg6KPB6C%2Fintegrations-manus-tile-en.webp?alt=media" alt="Manus AI tile opened with the request access button"><figcaption><p>Request access on the Manus AI tile.</p></figcaption></figure>

The walkthrough is in the FAQ: [Connect Manus AI to the Data Assistant](https://docs.adpage.io/faq/marketing-optimization/connect-manus-ai-to-the-data-assistant).

### Getting good answers

**Name the container and the period.** "Last month for the Dutch shop" beats "recently". With access to many containers, an unqualified question forces a guess.

**Ask for the reasoning, not just the number.** Asking which query it ran lets you check the answer, and the queries are often worth keeping.

**It can only see what is connected.** A question about Meta spend on a container with no Meta connection cannot be answered from thin air. The [integrations overview](https://docs.adpage.io/data-and-insights/marketing-integrations/integrations-overview.md) shows what each source brings in.

**Cross-check anything that surprises you.** These assistants are good at finding things and occasionally confident about the wrong ones. Where a finding would change a budget, verify it in Data Studio first.

### Troubleshooting

**Claude cannot reach the server.** Check the URL is exactly `https://mcp.adpage.io/mcp`, and that your client supports remote MCP servers — older versions do not.

**Sign-in loops or fails.** Sign in to AdPage in a browser first, then retry. An expired session is the usual cause.

**It cannot find your container.** Search by domain rather than by name. Names in AdPage are not always what you call the shop.

**Answers are empty for a platform you use.** That platform is not connected on that container. Check the Integrations page.

Still stuck? Contact <support@adpage.io> with the assistant you are using and what you asked.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.adpage.io/data-and-insights/marketing-integrations/ai-assistants-mcp.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
