name: google-chrome description: Use a headless Chrome browser to read websites or take screenshots.
Google Chrome (Headless)
Allows the agent to browse the live internet, read documentation, or verify UI via screenshots.
Usage
Use the browser.py script to interact with web pages.
Reading a Page (Default)
Converts the page content to Markdown for easy reading.
python app/skills/chrome/scripts/browser.py "https://example.com"
Taking a Screenshot
Useful for visual verification.
python app/skills/chrome/scripts/browser.py "https://example.com" --action screenshot --output "example.png"
Getting Raw Text
Useful for simple data extraction without formatting.
python app/skills/chrome/scripts/browser.py "https://example.com" --action text
When to use
- When you need to read documentation that is not in the knowledge base.
- When you need to search for current information (combine with
web_searchtool to find URLs first). - When you need to verify if a deployed website is working correctly.