- name
- playwright-commander
- displayName
- Playwright Commander
- description
- A skill to interact with web browsers using Playwright for advanced UI automation, analysis, and debugging.
- version
- 0.1.0
Playwright Commander Skill
Description
This skill provides agents with a powerful interface to programmatically control web browsers using Playwright. It enables advanced UI automation, web content analysis, and debugging capabilities beyond what the default browser tool might offer directly.
Core Functionalities
- Launch and close browser instances.
- Navigate to specified URLs.
- Capture screenshots of web pages.
- Retrieve HTML content or readable text from pages.
- Execute custom JavaScript within the browser context.
- Interact with web elements (click, type, fill forms) using various selectors.
- Retrieve element attributes or text content.
Usage
This skill is designed for programmatic interaction via exec commands, typically orchestrated by an agent.
# Example: Launch browser and navigate
python3 scripts/playwright_cli.py launch_browser --browser chrome --headless true
# Example: Navigate to URL and take screenshot
python3 scripts/playwright_cli.py navigate --url "https://example.com" --screenshot "path/to/screenshot.png"
# Example: Get page content
python3 scripts/playwright_cli.py get_content --url "https://example.com" --type html
# Example: Click an element
python3 scripts/playwright_cli.py click_element --url "https://example.com" --selector "button.submit"Prerequisites
- Playwright Python package (
playwright) installed in the skill's virtual environment. - Playwright browser binaries installed (
playwright install).
Configuration (config.json)
(To be defined if specific configuration beyond default Playwright options is needed.)
Purpose
To give me more granular and robust control over web browser interactions, enabling more complex UI automation, detailed web page analysis, and better troubleshooting capabilities for browser-related issues, especially for the Mac App conversion project.