Does this skill support Firefox?

Yes, it supports Firefox through the Marionette protocol. You should launch Firefox with the --marionette flag to allow the skill to connect.

Can I run custom JavaScript in the browser?

Yes, the 'run' command allows you to execute any valid JavaScript string directly within the context of the active browser tab.

Can I take screenshots of the browser?

Yes, the skill provides a screenshot command that captures the current view or the full page (on Chrome) and saves it to a designated artifacts folder.

What happens if I have orphaned browser processes?

The skill includes a 'cleanup' command that identifies and terminates browser instances running with remote debugging flags to free up system resources.

How do I start Chrome for use with this skill?

You must launch Chrome with the --remote-debugging-port=9222 and --user-data-dir flags. On macOS, it is recommended to use 'open -a' to avoid permission issues.

Browser Controller

Name: Browser Controller
Author: smykla-skalski

bysmykla-skalski

0•

Browser Automation

Controls Chrome and Firefox browsers programmatically to automate web navigation, DOM interaction, and data extraction.

Browser Controller provides a robust interface for Claude to interact directly with web browsers through Chrome DevTools Protocol (CDP) and Firefox Marionette. It enables the automation of complex web-based tasks such as filling forms, clicking elements, taking screenshots, and executing custom JavaScript on live browser instances. This skill is essential for developers performing web scraping, automated end-to-end testing, or research workflows that require real-time interaction with dynamic web applications and visual verification.

Key Features

010 GitHub stars

02Capture of screenshots and full-page renders for visual auditing

03Direct execution of JavaScript snippets within the browser context

04Comprehensive tab management, navigation, and active tab switching

05Programmatic control of Chrome and Firefox via CDP and Marionette protocols

06Automated DOM interaction including CSS selector-based clicking and form filling

Use Cases

01Automating multi-step web scraping workflows on JavaScript-heavy websites

02Gathering research data from web applications that require session-based interaction

03Executing automated UI tests and visual regression checks during development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add smykla-skalski/research browser-controller

For use in Claude.ai and ChatGPT

Key Features

010 GitHub stars

02Capture of screenshots and full-page renders for visual auditing

03Direct execution of JavaScript snippets within the browser context

04Comprehensive tab management, navigation, and active tab switching

05Programmatic control of Chrome and Firefox via CDP and Marionette protocols

06Automated DOM interaction including CSS selector-based clicking and form filling

Use Cases

01Automating multi-step web scraping workflows on JavaScript-heavy websites

02Gathering research data from web applications that require session-based interaction

03Executing automated UI tests and visual regression checks during development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add smykla-skalski/research browser-controller

For use in Claude.ai and ChatGPT