Browser Control

Enable OpenClaw to control web browsers for scraping, testing, and automated interactions.

Playwright Integration

Configure the browser automation engine:

browser:
  engine: "playwright"
  headless: true
  timeout: 30000
  viewport:
    width: 1920
    height: 1080
  user_agent: "Mozilla/5.0 (compatible; OpenClawBot/1.0)"

Web Scraping

Extract data from websites:

tools:
  web_scraper:
    enabled: true
    max_pages: 50
    rate_limit: 2  # requests per second
    respect_robots_txt: true

Example usage in chat:

User: Scrape product prices from example.com/products
Agent: *launches browser, navigates, extracts data*
Found 24 products. Price range: $10-$299

Screenshot Capture

Take screenshots of web pages:

tools:
  screenshot:
    format: "png"
    quality: 90
    full_page: true
    max_height: 10000

Use in conversations:

User: Screenshot https://example.com
Agent: *captures and uploads image*

Form Filling Automation

Automate form submissions:

tools:
  form_filler:
    enabled: true
    wait_for_navigation: true
    submit_timeout: 10000

Example workflow:

User: Fill login form at example.com with test credentials
Agent: *fills form, submits, returns result*

Security Considerations

• Always run browsers in headless mode in production
• Set appropriate timeouts to prevent hanging sessions
• Use respect_robots_txt: true for ethical scraping
• Restrict browser access to trusted domains when possible