Interactive Browser Mode

Interactive mode lets you take direct control of the agent's browser session. You can click, type, and navigate while the agent observes.

Switching to Interactive Mode

When the browser panel is visible during a session:

1. Click the Interactive toggle in the browser panel header
2. The browser switches from view-only to interactive
3. Your mouse clicks and keyboard input go directly to the browser

What You Can Do

In interactive mode:

• Click on any element in the browser
• Type text into input fields
• Scroll up, down, and sideways
• Navigate by clicking links or entering URLs

When to Use Interactive Mode

Interactive mode is useful when:

• The agent gets stuck on a CAPTCHA or verification step
• You need to manually authenticate with a service
• You want to show the agent what to click or where to navigate
• Debugging browser automation — see exactly what the agent sees

View-Only Mode

The default mode is view-only — you can see what the agent is doing but can't interact. The agent has full control. Switch between modes at any time.

Mobile Support

Interactive mode works on mobile devices with touch input:

• Tap to click
• On-screen keyboard for text input
• Pinch to zoom
• Swipe to scroll

Fullscreen

Click the fullscreen button to expand the browser panel to fill your screen. Useful for detailed work or when you need to see the full page layout.