> For the complete documentation index, see [llms.txt](https://support.sayprimer.com/primer/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://support.sayprimer.com/primer/primer-support-review/section-mcp/mcp.md).

# MCP

Connect Primer to Claude, ChatGPT or another MCP client to ask about your performance and to build, change and run your audiences.

> **Beta** — opened on request. Contact <support@sayprimer.com> to request access.

Primer's MCP server lets an assistant you already use — Claude, ChatGPT, a coding agent, or any client that speaks the Model Context Protocol — work with your Primer data in plain language, without leaving the tool you're in. Requests are answered by the same assistant as [Chat](https://support.sayprimer.com/primer/primer-support-review/section-chat/using-chat.md), so you can ask your assistant what Primer can do and let it guide you.

## Adding Primer as a connector

Primer's MCP server lives at:

```
https://mcp.sayprimer.com/mcp
```

In Claude (claude.ai or desktop), that goes in:

```
Settings → Connectors → Add custom connector
```

Other clients differ — in ChatGPT and in CLI or IDE agents, add a custom MCP server with the same URL, following your client's own instructions. Adding the connector opens Primer's sign-in page, where you approve the connection.

## What you can do over MCP

* **Read performance** — campaign and conversion performance, the visitors, leads and accounts behind it, and how Primer works.
* **Manage audiences** — list and describe your audiences, build one from a description of who you want to reach, refine or rename it, and add or remove destinations you've already connected.
* **Run audiences** — Primer never runs an audience on the request that asks for it. It first says what the run will do, and runs it only once you confirm.

When something can only be done in the Primer app, Primer says so and links you to the place to do it.

## What data the connection can reach

A connection is scoped to **one organization**: the one you signed in to. If you belong to more than one, remove the connector and add it again, signing in to the organization you want.

Answers come back as written summaries with their supporting numbers, not as lists of records: there's no export of audience members or revealed visitors over MCP.

## Data privacy

Primer receives only what each request carries: your question or instruction, any targeting details your assistant passes along to build an audience, and a one-line general reason for the request with the names of the model and the assistant app asking (those only to understand how Primer is used). It never receives the rest of your assistant's conversation, and your assistant never receives your Primer password.

Primer keeps your MCP conversations under your user so a follow-up can continue where the last request left off. They don't appear in your Chat history in the app. Remove the connector in your assistant to stop it reaching Primer.

Trouble getting an answer? See [An MCP question won't answer](https://support.sayprimer.com/primer/primer-support-review/section-troubleshooting/mcp-questions.md).


---

# 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://support.sayprimer.com/primer/primer-support-review/section-mcp/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.
