Automating the Browser
Drives headless browser interactions including clicking, typing, scrolling, and JavaScript execution before DOM extraction.
The Automating the Browser skill provides expert guidance for using Crawlberg's interactive CLI (`crawlberg interact`) to extract web content that requires scripted browser workflows. It details the exact JSON action schema for running sequential browser steps—such as clicking buttons, submitting forms, scrolling to trigger infinite load, waiting for elements, executing custom JavaScript, and taking screenshots—all while managing enforced limits and configuring external Chrome DevTools Protocol (CDP) endpoints.
Características Principales
01Built-in action limits and timeout enforcement for resilient scraping
02165 GitHub stars
03Sequential browser action scripting using strict JSON schemas
04Detailed JSON output capturing final DOM, screenshots, and per-action execution data
05Support for click, type, press, scroll, wait, executeJs, and screenshot actions
06External Chrome DevTools Protocol (CDP) connection support via WebSocket URLs
Casos de Uso
01Scraping dynamic web content hidden behind infinite scroll feeds or pagination buttons
02Capturing full-page screenshots and executing page-level JavaScript in headless environments
03Automating authentication workflows and multi-step form submissions before DOM extraction