selenium-browser

What This Skill Does

The selenium-browser skill is a powerful automation utility for OpenClaw that enables programmatic interaction with the web. By leveraging the Selenium framework, it launches a headless or visible Chrome instance to navigate to specific URLs, rendering pages as a standard browser would. Once the page is loaded, the skill captures a full screenshot, effectively transforming dynamic web content into a static visual file. This is particularly useful for verifying website layouts, monitoring page states, or creating archives of web data. The skill automatically handles the lifecycle of the browser, ensuring that resources are cleaned up once the task is complete.

Installation

To integrate this skill into your environment, use the OpenClaw hub command: clawhub install openclaw/skills/skills/andreybespalov89/selenium-browser

Ensure that you have google-chrome and the appropriate version of chromedriver installed on your host system. You can optionally configure their paths using the provided scripts/_env.sh file to match your specific system architecture.

Use Cases

• Visual Regression Testing: Confirm that your website UI renders correctly across different environments.
• Content Archiving: Take persistent snapshots of web articles, social media posts, or dashboards for offline analysis.
• Automated Reporting: Capture screenshots of internal BI tools or status pages to include in automated status reports.
• Web Reconnaissance: Safely view a page through a proxy to analyze its structure or content without revealing your primary IP address.

Example Prompts

1. "Open Chrome to https://www.google.com and take a screenshot for me."
2. "Can you use the browser skill to grab a screenshot of https://example-dashboard.local using the proxy http://proxy:3128?"
3. "Launch the selenium-browser for https://news.ycombinator.com in headless mode and tell me when it's done."

Tips & Limitations

• Wait Times: The script currently uses a static 5-second time.sleep(). For complex or slow-loading JavaScript applications, you may need to edit the Python script to utilize WebDriverWait for more robust element detection.
• Headless Mode: Always use the --headless flag when running the skill on server environments lacking a display server (X11).
• Resource Management: Selenium can be memory-intensive. Running multiple instances simultaneously may lead to system slowdowns or out-of-memory errors on smaller hardware nodes.