agent-browser

Control a real Chrome browser via Chrome DevTools Protocol for web automation.

12|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/CalebDane7/agent-browser --skill agent-browser-calebdane7
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/CalebDane7/agent-browser/tree/main
Command: npx skills add https://github.com/CalebDane7/agent-browser --skill agent-browser-calebdane7

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

AI agents often struggle to interact with live websites using standard automation tools that are detectable or require heavy dependencies. agent-browser lets AI agents drive your real Chrome via CDP, preserving your cookies, running headed, and avoiding redundant relays.

Core Features & Use Cases

  • Direct CDP-based control of your actual Chrome browser (headed or headless) with a single WebSocket.
  • Real-session isolation: separate cookies, localStorage, and tabs per agent/session; supports parallel tasks.
  • AI-native element references (@eN) and compact snapshots to dramatically reduce prompt size and improve reliability.
  • Open-world browsing tasks: login, form fill, navigation, data extraction, screenshot capture, network interception, and JavaScript execution on live sites.
  • Use cases include verifying deployments, testing dashboards (Google Ads, Stripe, Cloudflare), and automating repetitive browser tasks.

Quick Start

Install agent-browser, run start-chrome-debug to launch a persistent Chrome, then use agent-browser to open a site, snapshot the UI to obtain @e refs, and interact using those refs.

Frequently Asked Questions about agent-browser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate Chrome browser tasks using AI agents via CDP?

AI agents automate Chrome browser tasks via CDP by connecting to a persistent Chrome instance started with start-chrome-debug, using WebSocket to drive web interactions like form filling and data extraction. This approach preserves cookies and supports headed UI.

What is the best way to control a real Chrome browser for web automation without losing cookies?

The best way to control a real Chrome browser for web automation while preserving cookies is using direct CDP-based control via a single WebSocket. This method runs headed, maintains your actual session, and avoids redundant relays.

How do I reduce prompt size when extracting web data with AI browser automation?

Reduce prompt size in AI browser automation by using AI-native element references (@eN) and compact snapshots. This technique dramatically shrinks token usage and improves interaction reliability when extracting web data.

Can I run parallel browser automation tasks with isolated sessions in Chrome?

Yes, you can run parallel browser automation tasks with real-session isolation in Chrome. This setup separates cookies, localStorage, and tabs per agent or session, allowing multiple tasks to execute simultaneously without interference.

Does CDP-based browser automation work for testing dashboards like Google Ads and Stripe?

CDP-based browser automation works for testing dashboards like Google Ads, Stripe, and Cloudflare. It drives live web apps to verify deployments, fill forms, capture screenshots, and automate repetitive browser tasks across modern sites.

What are the limitations of using CDP for open-world browsing tasks?

A limitation of using CDP for open-world browsing tasks is that it requires a persistent Chrome instance started with start-chrome-debug. Without this specific environment setup, AI agents cannot connect to drive the browser or execute JavaScript on live sites.