chrome▌

Puppeteer-based browser automation with JSON output, screenshot compression, and performance analysis. \n \n Includes 10+ scripts for navigation, screenshots, clicking, form filling, JavaScript execution, element discovery, console monitoring, network tracking, and Core Web Vitals measurement \n Automatic screenshot compression using ImageMagick keeps files under 5MB for API compatibility; supports custom size thresholds and format options \n Supports command chaining by keeping browser sessions

Diagnose and resolve Claude in Chrome MCP extension connectivity conflicts on macOS. \n \n Addresses the primary issue: Claude.app (Cowork) and Claude Code CLI use competing native messaging hosts with incompatible socket formats; fix by disabling one config and restarting both Chrome and Claude Code \n Provides quick diagnosis commands to identify which native host is running, locate socket files, and verify active configurations \n Includes a toggle script to switch between Claude.app and Clau

Chrome DevTools integration for debugging, performance analysis, and browser automation via MCP. \n \n Provides tools for page navigation, element inspection via snapshots, clicking, form filling, screenshot capture, and JavaScript evaluation \n Operates on a persistent Chrome profile with automatic browser startup; switch between multiple pages using list_pages and select_page \n Supports efficient workflows: navigate, wait for content, take snapshots to identify elements by uid , then interact

Skill by ara.so — Daily 2026 Skills collection.

Automate browser tasks in the user's real Chrome session via the agent-browser CLI.

Standalone script: No download; the skill invokes .claude/tools/chrome-browser/chrome-browser.cjs (Node.js v18+ required).