Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Web scraping skill using Playwright with anti-bot bypass techniques. Supports three scraping modes: built-in fetch for simple sites, Playwright Simple for JavaScript-rendered pages, and Playwright Stealth for Cloudflare-protected sites. Stealth mode hides automation markers and uses realistic device user agents.
Playwright Stealth achieves 100% success on Cloudflare-protected sites where standard scrapers, Crawlee, and Puppeteer all fail.
When to use it
- Scraping Cloudflare-protected forums and community sites
- Extracting content from JavaScript-heavy single-page apps
- Automating data collection from sites that block standard scrapers
- Taking screenshots and saving HTML snapshots of dynamic pages
- Bypassing 403 errors when web_fetch returns blocked responses
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: scrape the hot posts from https://m.discuss.com.hk/#hot
- 1Identifies site as Cloudflare-protected based on domain
- 2Runs playwright-stealth.js with the target URL
- 3Script injects init scripts to hide navigator.webdriver and sets iPhone user agent
- 4Waits with randomized delays to mimic human browsing
- 5Returns JSON output with title and content, plus saves screenshot and HTML
JSON object with page title, extracted content, and elapsed time; screenshot and HTML saved to disk