agent-browser

Automate Chromium browser interactions via CLI for navigation, extraction, and testing.

Updated Feb 5, 2026
One-click install
npx skills add https://github.com/hungson175/shared-claude-config --skill agent-browser-hungson175
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/hungson175/shared-claude-config/tree/main/skills/agent-browser
Command: npx skills add https://github.com/hungson175/shared-claude-config --skill agent-browser-hungson175

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill automates the tedious and time-consuming task of controlling web browsers for AI agents, enabling automated navigation, interaction, data extraction, and testing from a single CLI.

Core Features & Use Cases

  • Headless browser automation for AI agents with accessible element targeting using semantic locators.
  • Multi-session isolation to run concurrent browsing tasks without cross-contamination.
  • Data extraction, form filling, testing, and web interaction workflows across websites.

Quick Start

Install the CLI and Chromium, then start automating tasks:

  • npm install -g agent-browser # Install the CLI
  • agent-browser install # Download Chromium
  • agent-browser open "https://example.com" # Open a page
  • agent-browser snapshot # Retrieve accessibility tree and element references
  • agent-browser find role button click --name "Submit" # Interact with elements
  • agent-browser screenshot output.png # Capture a visual state

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate web tasks using a headless browser from the command line?

You can automate web tasks by driving a Chromium-based browser through a CLI. This allows automated navigation, element interaction, data extraction, and testing from a single command line interface without a visible browser window.

How do I locate and interact with web elements during browser automation without standard CSS selectors?

You can interact with web elements using semantic locators. By taking an accessibility snapshot, you retrieve element references and target them by role and name, such as finding a button by its accessible name to click it.

What do I need to install to run Chromium-based browser automation for AI agents?

You need Node.js installed on your system. After installing the CLI package globally via npm, you must download the Chromium browser using the install command, with Linux users advised to use the --with-deps flag for full compatibility.

Can I run concurrent web scraping and testing tasks without sessions interfering with each other?

Yes, you can run concurrent browsing tasks using multi-session isolation. This feature prevents cross-contamination between sessions, allowing multiple web scraping and interaction workflows to operate simultaneously in the same browser environment.

What is the best way to extract data from dynamic web pages using CLI browser automation?

The best way to extract data from dynamic pages is to use a headless browser CLI to navigate the URL, take an accessibility snapshot of the rendered DOM, and then use commands to interact with and pull data from the targeted elements.