Playwright (Automation + MCP + Scraper)

What This Skill Does

The Playwright skill for OpenClaw is a powerful browser automation and web scraping framework designed to handle modern, JavaScript-heavy web applications with ease. Unlike basic HTML fetchers, this skill utilizes a real headless browser engine, enabling it to render dynamic content, execute complex interactions, and navigate multi-page workflows. It serves as an essential tool for developers and data engineers who need to convert web interactions into repeatable, automated tasks.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface: clawhub install openclaw/skills/skills/ivangdavila/playwright Ensure your local environment has the necessary system dependencies for Playwright, such as the required browser binaries. You may need to run the installation script provided by the playwright package to ensure full browser support.

Use Cases

1. Web Scraping: Extract data from dynamic tables, infinite scrolling lists, or pages that require JavaScript execution to populate content.
2. Automated Form Submission: Streamline repetitive data entry tasks, automated logins, or complex checkout processes in testing environments.
3. End-to-End (E2E) Testing: Build robust verification suites for web applications, ensuring UI components behave as expected across different navigation flows.
4. Media Capture: Automatically generate high-resolution screenshots or high-fidelity PDFs of dynamic dashboard states for reporting.
5. Workflow Automation: Bridge the gap between disparate web services that do not offer APIs by interacting directly with their front-end interfaces.

Example Prompts

1. "Open the dashboard at example.com, login using the saved session, and extract all rows from the 'Pending Tasks' table into a CSV format."
2. "Navigate to the product page, wait for the dynamic price to load, take a full-page screenshot, and save it to the local exports folder."
3. "Fill out the contact form on example.com/support using the provided user credentials, submit it, and verify that the success message appears."

Tips & Limitations

• Priority Selectors: Always prioritize getByRole or getByLabel over CSS classes to ensure your automation remains resilient against UI updates.
• Performance: Avoid waitForTimeout() as it slows down your agent. Use explicit wait conditions like waitForURL or waitForSelector to improve reliability.
• Memory Management: Always wrap your execution in a try/finally block to ensure browser.close() is called, preventing memory leaks during long-running tasks.
• Security: Be mindful when providing credentials to the agent. Utilize the storageState feature to persist authentication sessions rather than re-entering credentials for every execution loop.