UXMachine for agents

Your coding agent asks UXMachine to measure your site. We open the page in a real browser, on our infrastructure, and send back what we observed: what, where, with which measurement, and how far that measurement goes. What to do about it is up to you, or your agent. Then you measure again, until it is no longer observed.

What each observation contains

What is observed and where: the page, the window width and the element, with the measurement it comes from.
The observable state to bring it to.
How to check that it got there: which measurement runs again and which value counts as reached.
If we cannot state something, we say so and say why.

Measuring again

When you measure again, each observation comes back with one of these four states: "Still observed", "No longer observed", "New" or "Not comparable", with its reason.

What it measures and what it does not

We measure what the page shows when it loads, in two windows (1280×800 and 390×844) and on up to 11 pages of the site. Only checks with an exact state that can be measured again the same way come out through this channel; they are on this list: /en/agents/rules. We do not fill in forms, we do not enter password-protected areas and we do not follow what happens after a click. We do not give opinions on design, copy or brand. This does not replace an audit or testing with people.

Installation

Add the UXMachine connector to your tool with this address: https://uxmachine.app/mcp

The first time, your tool opens a UXMachine window so you can authorize access with your account. If you do not have an account, you can create one there.

Ask your agent to measure a page. For example: "Measure the home page with UXMachine."

Lovable

In your Lovable project, open Connectors → Add MCP server and enter this address: https://uxmachine.app/mcp

In the UXMachine window, sign in (or create your account) and click "Authorize".

In the project chat, ask: "Measure this project's home page with UXMachine." Lovable already knows which address to measure.

Claude Code

In your terminal, run: claude mcp add --transport http uxmachine https://uxmachine.app/mcp

Inside Claude Code, type /mcp and choose uxmachine.

Click "Authenticate". The UXMachine window opens in your browser.

Click "Authorize". Go back to Claude Code and ask for a measurement.

Cursor

In Cursor, open Customize → MCPs → New MCP Server and paste this into mcp.json: {"mcpServers":{"uxmachine":{"url":"https://uxmachine.app/mcp"}}}

Click "Authenticate". In the UXMachine window, click "Authorize" and then "Open Cursor".

In Cursor's chat, use Agent mode.

When Cursor asks for permission to use medir and estado, click "Run".

On every pull request (GitHub Action)

If your project is on GitHub and each pull request has its own preview, the Action measures that preview and leaves a comment with what changed compared to the base branch. It works with whatever tool you build your site with. It uses a key from your account, which you create in your dashboard and store as a secret in the repository.

robots.txt and other people's sites

The channel respects robots.txt: if a site does not allow us to visit it, we do not measure it and we tell you so. Only ask for measurements of sites that are yours or that you have permission to review.

Your local version

Your agent requests a tunnel with the tunel tool and runs the command it receives on your machine; it only needs Node. The tunnel closes on its own after 30 minutes. The measurement runs on our infrastructure, through that tunnel.
Measurements made through the tunnel or from the GitHub Action are marked as such.

Usage limits

One measurement at a time per account and up to 10 a day. Each measurement uses 1 credit from your account.

What we keep when you connect an agent: /en/privacy#agent-channel