Can I use my existing logged-in accounts and cookies?

Yes, by using the --profile flag when starting the browser, the skill mirrors your actual Chrome profile so you stay logged in to your services.

Can I execute complex, multi-line JavaScript?

Yes, though you should wrap multiple statements in an IIFE (Immediately Invoked Function Expression) and ensure the string is passed as a single line.

Does this skill require a separate Chrome installation?

Yes, it utilizes your local Chrome installation and connects via the remote debugging port (defaulting to 9222).

How does the 'pick' tool work?

The pick tool opens an interactive mode where you can visually click elements in the browser window to return their selectors or text directly to Claude.

Is this a replacement for the Model Context Protocol (MCP)?

It serves as a lightweight alternative to MCP specifically for browser tasks, focusing on direct script-based interaction with the Chrome DevTools Protocol.

Browser Automation & Scraping

Name: Browser Automation & Scraping
Author: iamzhihuix

byiamzhihuix

•

232

浏览器自动化

Automates Chrome via the DevTools Protocol to navigate pages, execute JavaScript, and extract web data interactively.

关于

This skill provides a lightweight suite of tools leveraging the Chrome DevTools Protocol (CDP) to enable seamless browser automation and web scraping directly within Claude Code. It allows users to launch Chrome with custom or existing profiles to maintain authenticated sessions, interactively select DOM elements, execute asynchronous JavaScript, and capture screenshots for visual verification. It is an ideal solution for developers who need to scrape dynamic content or automate web-based workflows without the complexity of a full Model Context Protocol (MCP) implementation.

主要功能

Profile mirroring to maintain authenticated sessions and logged-in states
Asynchronous JavaScript execution directly in the page context
232 GitHub stars
Interactive DOM element selection with visual picking functionality
CDP-based Chrome automation for high-performance browser control
Automated screenshot capture for visual debugging and site exploration

使用场景

Automating repetitive web tasks behind login walls using local profiles
Extracting precise DOM selectors and JSON payloads for development and testing
Scraping data from dynamic, JavaScript-heavy websites and SPAs

关于

主要功能

Profile mirroring to maintain authenticated sessions and logged-in states
Asynchronous JavaScript execution directly in the page context
232 GitHub stars
Interactive DOM element selection with visual picking functionality
CDP-based Chrome automation for high-performance browser control
Automated screenshot capture for visual debugging and site exploration

使用场景

Automating repetitive web tasks behind login walls using local profiles
Extracting precise DOM selectors and JSON payloads for development and testing
Scraping data from dynamic, JavaScript-heavy websites and SPAs