Use the built-in browser

The app has a browser built into it. You can open pages in it yourself, and the agent can use it to do tasks on the web for you.

Open a page

Open the browser panel and type an address in the field at the top.

The panel works like a small browser: it has tabs, and buttons to go back, go forward and reload.

Let the agent use it

An empty workspace suggests automating a repetitive web task as one of its starting points. The agent drives the same browser panel you see.

Because the panel is visible, you can watch what the agent does and stop it.

Note

The browser panel only exists in the desktop app.