browser-automation

What This Skill Does

The browser-automation skill is a powerful utility for managing web-based tasks programmatically. Unlike standard testing frameworks designed for regression, this skill is optimized for production-grade browser orchestration. It handles complex tasks such as navigating dynamic websites, executing authenticated sessions, scraping data into structured formats, and automating repetitive form submissions. By leveraging industry-standard tools like Playwright, Puppeteer, and Selenium, the skill provides a robust headless browser environment capable of executing JavaScript, handling cookies, and managing DOM interactions as if a human user were present. It is designed to bridge the gap between simple API requests and manual navigation.

Installation

To integrate this skill into your environment, use the OpenClaw CLI tool. Run the following command in your terminal: clawhub install openclaw/skills/skills/anton-abyzov/sw-browser-automation Ensure that you have Node.js and the necessary runtime dependencies configured, as browser drivers may require specific environment settings for headless execution.

Use Cases

• Web Scraping: Extracting product pricing, stock availability, or blog content from sites that render content dynamically via React or Vue.js.
• Workflow Automation: Automating repetitive data entry into legacy enterprise web portals that lack public APIs.
• Asset Generation: Automatically navigating to reports or dashboards to capture high-fidelity screenshots or generate PDF exports for status reporting.
• Session Management: Handling multi-step login flows, including 2FA bypass (if configured) and cookie persistence for long-running monitoring tasks.

Example Prompts

1. "Open the URL https://e-commerce-site.com, scrape the top 20 items in the 'Electronics' category including prices and links, and save the data to a JSON file."
2. "Navigate to the CRM login page, sign in with my credentials, fill out the support ticket form with the details provided in this message, and submit it."
3. "Take a full-page screenshot of the dashboard at https://analytics.internal-portal.com after ensuring all charts have finished loading."

Tips & Limitations

• Rate Limiting: When scraping, always incorporate delays between requests. Respect robots.txt guidelines to avoid being blacklisted by target servers.
• Dynamic Elements: Always use wait strategies like waitForSelector or networkidle rather than static timeouts to ensure your script is resilient to fluctuating network speeds.
• Headless Mode: Run in headless mode by default for performance, but switch to visible mode (headed) for debugging and visual inspection of your automation scripts.
• Disclaimer: This tool is for automation, not testing. For UI/UX testing or end-to-end testing suites, please utilize the dedicated e2e-playwright skill instead.