Introducing Heap MCP: your Heap data, wherever you ask questions

Heap MCP gives you a new way to ask questions about your Heap data from the AI tools you already work in.

No new dashboards to learn. No waiting on an analyst. Just ask in plain language and get an answer grounded in your own Heap data.

For the first time, you can reach Heap's analytics engine from outside Heap. Here's what that means for your team, how it works, and how to get connected in a few minutes.

What is Heap MCP?

Heap MCP is built on the Model Context Protocol (MCP), an open standard introduced by Anthropic. MCP enables AI assistants and agents to connect directly to the data and tools behind your products.

With Heap MCP, you can connect your AI assistant of choice to your Heap project securely, while also understanding your events, properties and segments and running analyses on your behalf. You ask the question. Heap has the data. Your agent provides the analysis. The answer shows up right where you're working.

Better insights, for everyone, even faster

Great analytics shouldn't be reserved for the people who know where every chart lives. Heap MCP puts answers within reach of the whole team:

• Product managers can check feature adoption before a roadmap review, without filing a request.

• Marketers can see how a campaign performed and what those visitors did next.

• Designers and researchers can back up a hunch with real user behavior in the middle of their workflow.

• Engineers can pull usage data into the same place they're already building.

Because Heap MCP works inside your AI tool of choice, you can also bring Heap data together with the other tools you've connected, such as your data warehouse, CRM, user research repository or error monitoring. Ask one question across all of them. No more exporting, uploading and stitching it together by hand.

How it works

Getting from question to insight takes three steps:

• Connect. Add Heap MCP to your AI tool and sign in with your Heap account. You approve what it can access.

• Ask. Type your question the way you'd ask a teammate: "How did sign-ups trend over the last 30 days compared to the previous period?"

• Get your answer. Heap MCP finds the events, segments and properties that match your question, runs the analysis, and sends back the results for your AI assistant to explain.

Heap MCP is read-only. It can look at your data, but it can't create, change or delete anything in your project.

Works with the tools you already use

Heap MCP works with any AI tool that supports MCP, including:

• Claude (desktop, web, mobile and Claude Code)

• Cursor

• ChatGPT

• VS Code, and other MCP-compatible tools

See it in action: Using Heap MCP

Click here to see a full demo

How to get started

You'll need:

• A Heap account with access to Heap MCP

• A supported LLM plan that supports MCP 

Step 1: Open your LLMs connector settings

Step 2: Look for “Heap” in the available connectors

Step 3: Click connect” and  the Heap MCP details

Can’t add or don’t see it? An organization admin may need to add the connector first. After that, individual members can connect it from their own settings.

Step 4: Sign in to Heap

You will be prompted to connect or approve permissions. Sign in with your Heap account and approve access.

Step 5: Choose your environment

If your Heap account has more than one environment, tell your LLM which one to use, for example: "Use my Production environment." Heap remembers your choice.

Step 6: Ask your first question

Start a new chat, make sure the Heap connector is turned on, and ask away.

A few things to know

We built Heap MCP to be useful from day one, and to be careful with your data:

• Read-only. Heap MCP can't create, edit or delete anything in your project.

• Privacy-first. Only aggregated data and metadata are returned. No emails, names or other personal identifiers. Your data isn't used to train third-party AI models.

• Preset date ranges. Choose from today, yesterday, the last 7, 30 or 90 days, and the year, with comparisons to the previous period. Custom and all-time ranges are on our roadmap.

• New analyses only. Heap MCP runs fresh, one-off analyses. Saved charts and dashboards, and session replays, aren't included yet.

Get started

Heap MCP is available now. Connect it to your LLM in a few minutes, ask your first question, and see how much faster better insights can get.