Now, entire workflows happen in AI tools like Claude, ChatGPT, and Cursor. They’re becoming the places where people brainstorm, write code, and make decisions. Our own teams spend countless hours in these tools every week, and that’s exactly why we moved quickly to develop an MCP server.

Pendo is now available as an MCP (Model Context Protocol) server, easily accessible in Claude, ChatGPT, and Cursor. That means you can connect your Pendo data to the AI tools you already use, and start asking questions, running analysis, and generating outputs without ever opening a new tab.

What is MCP, and why does it matter?

Model Context Protocol (MCP) is an emerging standard that allows AI agents to securely connect to external systems. Think of it as a secure bridge that connects your LLM to your external tools.

Instead of manually pasting context into prompts, your AI can now:

  • Query product usage data in real time
  • Analyze feature adoption trends
  • Pull NPS or feedback themes
  • Surface churn signals
  • Reference actual account-level behavior

And because MCP respects your existing Pendo permissions, users only access data they’re already authorized to see.

5 things you can do with MCP 

  1. Ask questions without leaving your workflow. You're deep in a Claude conversation thinking through a product decision. Instead of switching to Pendo to check adoption numbers, you just ask, "What's the adoption rate for this feature over the last 90 days?", and get the answer right there.
  2. Bring product data into your code editor. For teams using Cursor, Pendo data becomes part of the development context. Engineers can query usage patterns, check which features are seeing high error rates, or understand how a recent release has performed, all while staying in the flow of building.
  3. Power autonomous agents with behavioral data. Create agents that can speak to each other and accomplish tasks without human intervention. For example, you can create a churn risk agent that can automatically spot drop-off via Pendo MCP, trigger a workflow in Salesforce, and then send updates to the rep in Slack. 
  4. Run analysis conversationally. Instead of navigating Pendo's UI to build a report, describe what you want to understand and let the AI do the legwork. "Which accounts have dropped off in the last 30 days and what were they doing before they churned?" is a question you can now just ask in plain language and get instant data-backed results.
  5. Power smarter guides with behavioral intent. MCP gives Pendo Guides real-time context about what a user is trying to do. Instead of triggering guides based on simple page rules, agents can now read behavioral signals (where the user has been, what they've clicked, where they're stuck) and surface exactly the right guide at exactly the right moment. For example, if a user keeps navigating to a feature but never completes a key action, an agent can detect that intent and automatically trigger a contextual walkthrough — no manual rule-building required.

Why this is a big deal

We believe product intelligence should travel with you, not live behind a login in a single product. The teams doing the best work aren't the ones who always go back to the source. They're the ones who have the right data available at the moment of decision, wherever that moment happens to occur.

Pendo as an MCP server is a step toward that future. Your data is no longer confined to Pendo's UI. It's available in the tools where your team already thinks and works: making every conversation a little smarter, every output a little more grounded, and every decision a little more confident. 

Today, products are shipped in days. Decisions need to be made fast. Data needs to be accurate. Ideas need to be fresh. You shouldn’t have to settle for tools that require hours of analysis. And with Pendo, you don’t have to. Get access to your data whenever and wherever your work actually gets done.

What’s coming next?

We're just scratching the surface of what's possible with MCP. We're expanding our partner ecosystem to bring Pendo data into even more of the tools your teams rely on. Coming soon: integrations with Fin (Intercom's AI support agent), Miro (collaborative diagramming and product planning), and Planview (enterprise portfolio and work management), so your product intelligence can flow directly into customer conversations, whiteboard sessions, and strategic planning workflows.

Get started today

Pendo MCP is generally available in the Claude and Cursor directories. Connecting takes less than a minute: just search for Pendo in the directory, authenticate your account, and you're ready to start querying.

Pendo MCP respects your existing Pendo permissions, ensuring AI agents only access data users are already authorized to see.

Already set up with Pendo MCP? Hit the ground running with 50+ MCP prompts.