---
title: "How we connected Notion AI to 160 new apps in three weeks"
description: "Notion used the Model Context Protocol (MCP), an open standard for connecting AI tools to other apps, to expand its AI integrations. Previously, each new MCP connection required hours of manual setup, including creating accounts, navigating developer settings, and testing, so only 25 apps were connected. In three weeks, Notion grew that number to 185 by building a Notion agent that automates the pipeline from setup to production. The agent handles sign-ups, browser clicks, and tests, while humans still review code changes, design assets, and partner with vendors. A key lesson was that private developer apps could pass internal tests but fail for customers, so the agent now tests from a second account simulating real users."
---

# How we connected Notion AI to 160 new apps in three weeks

[Notion](https://yomu.fyi/company/notion) · Eileen Jiang · Oct 16, 2026

**Type:** Problem & solution

## Summary

Notion used the Model Context Protocol (MCP), an open standard for connecting AI tools to other apps, to expand its AI integrations. Previously, each new MCP connection required hours of manual setup, including creating accounts, navigating developer settings, and testing, so only 25 apps were connected. In three weeks, Notion grew that number to 185 by building a Notion agent that automates the pipeline from setup to production. The agent handles sign-ups, browser clicks, and tests, while humans still review code changes, design assets, and partner with vendors. A key lesson was that private developer apps could pass internal tests but fail for customers, so the agent now tests from a second account simulating real users.

## Approach / What changed

We built a Notion agent that runs each MCP server through a pipeline from setup to production release, automating sign-ups, browser interactions, and test execution. Humans still review code changes, design logos, and handle partner communication. The agent uses a Kanban board in Notion to track each MCP’s status and moves servers across columns, checking hourly for any eligible to advance.

## Takeaways

- Building an internal agent to automate account creation, CAPTCHAs, and browser-based setup can slash MCP integration time from hours to minutes.
- Private developer apps can cause MCP connections to pass internal tests but fail for customers; simulating real-user credentials in a second test account catches this.
- Splitting an agent runbook into one page per pipeline step prevents it from skipping steps due to excessive context in a single instruction.

## FAQ

### How does the agent handle waiting steps?

Some steps, like waiting for logo deployment or releasing during business hours, are asynchronous. The agent checks every hour to see if any MCPs are eligible to move forward.

### How did they avoid picking the wrong MCP?

The agent confirms the server URL on the vendor's website and on mcpservers.org, and a person approves every new MCP before it enters the pipeline to avoid confusion from similar names.

**Tags:** [Agents](https://yomu.fyi/topic/agents), [AI](https://yomu.fyi/topic/ai), [Automation](https://yomu.fyi/topic/automation), [MCP](https://yomu.fyi/topic/mcp), [Notion](https://yomu.fyi/topic/notion)

- Source: [Notion](https://www.notion.com/en-gb/blog/how-we-connected-notion-ai-to-160-new-apps-in-three-weeks)
- Source URL: https://www.notion.com/en-gb/blog/how-we-connected-notion-ai-to-160-new-apps-in-three-weeks
- Ingested by Yomu: 2026-10-16T18:00:19.689Z

[Read original post](https://www.notion.com/en-gb/blog/how-we-connected-notion-ai-to-160-new-apps-in-three-weeks)