Skip to main content

Connect Uxi using MCP

Learn how to connect Uxi with third-party services like n8n, Claude, ChatGPT, Gemini, Make.com through the MCP endpoint.

Vasil Dachev avatar
Written by Vasil Dachev
Updated this week

What is the MCP endpoint integration?

We’ve introduced a new way to integrate Uxi, our AI engine, with third-party services such as n8n, Claude, Make.com, Gemini, and ChatGPT. Using the MCP (Model Context Protocol) endpoint, these tools can securely pull data from your Uxify account and communicate directly with Uxi.

This opens up powerful automation and intelligence possibilities - from triggering workflows in automation platforms like n8n or Make.com, to extending Uxi’s real-time insights into AI agents like Claude, Gemini, or ChatGPT.

How it works?

The MCP endpoint acts as a bridge between your Uxify account and external services. Each Uxify account is assigned a unique API key. You’ll use this key when constructing your MCP endpoint URL.

The endpoint format looks like this:

https://mcp.uxify.com/?apikey=your-unique-api-key
  • Replace your-unique-api-key with your actual Uxify API key.

  • Keep this endpoint private - treat it like a secure key.

Example use cases

n8n

Automate complex workflows that include Uxi intelligence.
Example: When a performance regression is detected, n8n can call Uxi for details, then automatically create a Jira ticket or send a Slack alert.

Make.com

Use Make.com (formerly Integromat) to connect Uxi to marketing or analytics tools.
Example: When Uxi identifies high exit rates, Make.com can sync that insight with Google Sheets or trigger an email via HubSpot.

Claude

Let Claude query Uxi during a conversation.
Example: You can ask Claude, “What’s our current INP trend for checkout pages?” — and Claude will fetch that insight through the MCP endpoint.

Gemini

Extend Google’s Gemini with Uxify data.
Example: Gemini can generate reports or recommendations using live performance metrics from Uxi.

ChatGPT

Integrate Uxi into custom GPTs or GPT Actions (via API calls).
Example: Create a “Performance Coach” GPT that answers questions like, “Which pages have the worst CLS last week?” by calling Uxi’s MCP endpoint.

Things to keep in mind

  • Your API key is unique and private - treat it like a password.

  • Keep your MCP URL safe. Do not share it in public repositories or unsecured environments.

  • If you need to rotate or reset your access, contact our support team.

Next steps

  1. Copy your MCP endpoint URL.

  2. Paste it into the third-party tool (e.g., n8n, Make.com, Claude, Gemini, or ChatGPT) when asked for a connection or endpoint.

  3. Start building workflows and conversations that leverage Uxi.

Did this answer your question?