How do element references work?

When you take a snapshot, the tool labels interactive elements with IDs like @e1. You can then tell Claude to 'click @e1' which is more reliable than using long CSS selectors.

Does it handle dynamic pages and SPAs?

Yes, it includes advanced 'wait' commands that can trigger based on specific selectors, text appearances, URL patterns, or network idle states.

What tools are included in this browser skill?

It provides a full suite of MCP tools including browser/open, browser/snapshot, browser/click, browser/fill, browser/screenshot, and browser/close.

Can I use this for multi-agent workflows?

Yes, it supports isolated sessions using the --session flag, allowing multiple agents to work on different URLs or maintain separate authentication states simultaneously.

How does this skill reduce context usage for the AI?

It generates AI-optimized snapshots using accessibility trees and shorthand element references (@e1, @e2) instead of the full DOM, reducing token consumption by up to 93%.

Browser Automation & Web Intelligence

Name: Browser Automation & Web Intelligence
Author: ruvnet

byruvnet

•

12,774

•

Automatización de Navegadores

Automates web navigation and data extraction using AI-optimized accessibility snapshots that reduce token usage by 93%.

This skill empowers Claude to navigate the live web, interact with complex UI elements, and extract structured data using a specialized browser automation engine. By utilizing AI-optimized snapshots and element references (@e1, @e2) instead of raw DOM trees, it drastically reduces context consumption while maintaining high precision for tasks like automated research, form submission, and end-to-end testing. It integrates seamlessly with the Claude Flow orchestration platform, supporting multi-session workflows and isolated agent environments.

Características Principales

01Multi-session support for parallel agent swarm workflows

02Full MCP tool suite for clicking, filling, and screenshotting

03AI-optimized snapshots with 93% context reduction

04Advanced wait states for dynamic content and network idle

05Interactive element referencing for reliable, deterministic control

0612,774 GitHub stars

Casos de Uso

01End-to-end UI testing and regression checking during development

02Automated web scraping and data collection from dynamic applications

03Complex multi-step form completion and navigation for research tasks

Características Principales

01Multi-session support for parallel agent swarm workflows

02Full MCP tool suite for clicking, filling, and screenshotting

03AI-optimized snapshots with 93% context reduction

04Advanced wait states for dynamic content and network idle

05Interactive element referencing for reliable, deterministic control

0612,774 GitHub stars

Casos de Uso

01End-to-end UI testing and regression checking during development

02Automated web scraping and data collection from dynamic applications

03Complex multi-step form completion and navigation for research tasks