Does this skill require me to manually open Chrome?

No, the browser starts automatically on the first tool call using a persistent profile configured in your MCP server settings.

Can I see what the browser is doing visually?

Yes, you can use the take_screenshot tool at any time to capture the current visual state of the browser and inspect the UI.

How does Claude interact with web elements in this skill?

Claude uses the take_snapshot tool to generate a text-based representation of the page where every element is assigned a unique UID, allowing for precise clicks and text entry.

Can I run custom JavaScript through this skill?

Yes, the evaluate_script tool allows you to run custom JavaScript to retrieve data or trigger actions not covered by the standard accessibility tree.

What should I do if an element is not found by the AI?

You should instruct the skill to take a fresh snapshot, as the page state may have changed or the element may have been dynamically removed.

Chrome DevTools

Name: Chrome DevTools
Author: ChromeDevTools

byChromeDevTools

•

23,073

Automatización de Navegadores

Automates browser interactions, debugs web pages, and inspects network traffic directly within Claude.

Acerca de

This skill integrates Chrome DevTools with Claude via the Model Context Protocol (MCP), enabling sophisticated AI-driven browser automation and web debugging. It allows Claude to navigate web pages, interact with DOM elements using unique UIDs, capture visual screenshots, and evaluate custom scripts in real-time. By providing a persistent Chrome profile and deep inspection capabilities, it streamlines the process of troubleshooting frontend issues, automating repetitive web tasks, and analyzing web application performance without leaving the terminal.

Características Principales

Precision element interaction using unique UIDs and snapshot technology
23,073 GitHub stars
Multi-page management and context switching via list and select tools
Visual debugging through high-resolution page screenshots and traces
Automated browser lifecycle management with persistent Chrome profiles
Real-time script evaluation for advanced DOM and accessibility data

Casos de Uso

Debugging complex frontend layouts and JavaScript execution errors
Extracting structured data and analyzing network requests for performance
Automating end-to-end user workflows and browser-based testing

Acerca de

Características Principales

Precision element interaction using unique UIDs and snapshot technology
23,073 GitHub stars
Multi-page management and context switching via list and select tools
Visual debugging through high-resolution page screenshots and traces
Automated browser lifecycle management with persistent Chrome profiles
Real-time script evaluation for advanced DOM and accessibility data

Casos de Uso

Debugging complex frontend layouts and JavaScript execution errors
Extracting structured data and analyzing network requests for performance
Automating end-to-end user workflows and browser-based testing