Browser Automation V2

What This Skill Does

Browser Automation V2 is an enterprise-grade utility designed for the OpenClaw agent ecosystem, enabling sophisticated interaction with the web. Unlike basic scrapers, this skill implements robust resource management, ensuring that browser instances are cleaned up automatically after execution to prevent memory leaks and system bloat. It features smart waiting mechanisms, exponential backoff for retries during network volatility, and strict concurrency locks that prevent multiple agent tasks from conflicting over browser profiles. It is built to be modular, providing specific scripts for Google searching, batch URL processing, web content summarization, and automated form interaction.

Installation

To integrate this skill into your local OpenClaw environment, ensure you are running OpenClaw v2026.2.15 or later. You can install it directly via the CLI using the command: clawhub install openclaw/skills/skills/bodafu/browser-automation-v2. Once installed, you can configure your experience by exporting environment variables such as BROWSER_TIMEOUT and BROWSER_PROFILE. The skill will register itself into your workspace, allowing you to trigger tasks immediately using the run command within your agent chat interface.

Use Cases

This tool is ideal for developers and power users who need to automate complex web workflows. Use it to perform high-fidelity research by searching Google and capturing screenshots or PDFs of relevant documentation. It is also perfect for data gathering pipelines that require fetching dynamic content from multiple URLs in a batch. Furthermore, the fill-form.js component makes it invaluable for automated testing, onboarding scripts, or interacting with legacy web interfaces that lack modern APIs.

Example Prompts

1. "run search-google.js "latest updates on OpenClaw API version 2026" and save the results as a PDF."
2. "Please process these URLs and summarize the main content for me: https://example-blog.com/post1, https://example-docs.com/guide."
3. "Fill out the contact form at https://support.example.com using the email [email protected] and the message 'Need help with browser automation'."

Tips & Limitations

To get the most out of this skill, monitor the DEBUG logs if tasks fail; setting DEBUG=1 will expose the internal logic of the browser interactions. If you encounter frequent timeout errors, consider increasing the BROWSER_TIMEOUT variable to account for slow page loads. Note that the skill relies on browser selectors; if the target website changes its structure frequently, you may need to update your reference selectors. Always ensure your browser profile is not locked by another running instance before initiating heavy batch jobs to avoid potential race conditions.