Browser Automation: Forms, Screenshots, and More

Browser Automation

OpenClaw's browser tool goes beyond simple web reading. It controls a real browser (Chromium) to navigate websites, interact with elements, and automate complex web tasks.

Prerequisites

• Chromium or Chrome installed
• Browser tool enabled in configuration

{
  "tools": {
    "browser": {
      "enabled": true,
      "headless": true
    }
  }
}

Capabilities

Navigation

• Open URLs and follow links
• Navigate multi-page flows
• Handle redirects and popups

Interaction

• Click buttons and links
• Fill form fields
• Select dropdown options
• Upload files
• Handle checkboxes and radio buttons

Capture

• Take full-page screenshots
• Capture specific elements
• Extract text from rendered pages
• Save pages as PDF

Sessions

• Login to websites
• Maintain authenticated sessions
• Handle cookies and tokens

Example Tasks

Price Comparison

"Compare the price of AirPods Pro on Amazon, Best Buy, and the Apple Store. Take a screenshot of each listing."

Form Submission

"Go to the feedback form at example.com/feedback, fill in my name, email, and the message 'Great service!', then submit."

Content Monitoring

"Check if the job posting at company.com/careers/senior-dev is still active. If it is, save the full description."

Data Extraction

"Go to the weather service website and get the 5-day forecast for San Francisco. Format it as a table."

Headless vs Visible

• Headless (default): Browser runs invisibly in the background
• Visible: Browser window appears so you can watch the automation

{
  "tools": {
    "browser": {
      "headless": false
    }
  }
}

Security

• Browser sessions are isolated from your personal browser
• No access to your saved passwords or bookmarks
• Each session starts fresh unless explicitly configured otherwise
• Be cautious about which sites you automate — respect Terms of Service