Browser use

Agent can open a browser, navigate web pages, click elements, fill forms, and take screenshots. This is useful for testing frontend changes and debugging UI issues.

Can Cursor control my browser?

Yes. Agent can launch a browser, navigate to any page, interact with elements, and take screenshots. Use it to test UI changes, verify form behavior, or debug visual issues without leaving Cursor.

How can I get Agent to use the browser?

Ask Agent to open a browser and it handles the rest. For example: "Open the signup page and check that form validation works."

Agent can:

• Navigate to URLs
• Click buttons and links
• Fill in form fields
• Take screenshots
• Read page content

What sites work with browser use?

Agent can browse local development servers and public websites. For pages behind a login, tell Agent how to sign in.

Related

• Browser reference