agent-browser

Automate browser tasks via the agent-browser CLI with snapshot-based refs.

2|Updated Sep 28, 2025
One-click install
npx skills add https://github.com/abhay-rana/create-modern-react --skill agent-browser-abhay-rana
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/abhay-rana/create-modern-react/tree/main/.claude/skills/agent-browser
Command: npx skills add https://github.com/abhay-rana/create-modern-react --skill agent-browser-abhay-rana

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill enables automated control of web browsers via the agent-browser CLI, turning repetitive web interactions into repeatable, scriptable tasks.

Core Features & Use Cases

  • Single-command browser automation: Open URLs, snapshot DOM to get compact refs, fill forms, click controls, take screenshots, and extract page data with minimal AI input.
  • Multi-session and state persistence: Run isolated sessions with independent cookies, storage, and tabs; optionally save and restore session state for CI/CD or testing pipelines.
  • Use Case: Automate end-to-end form submissions, perform UI testing, capture site content for documentation, and generate reproducible walkthroughs.

Quick Start

  • agent-browser open https://example.com
  • agent-browser snapshot -i
  • agent-browser fill @e1 "sample"
  • agent-browser click @e1
  • agent-browser screenshot ./page.png

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate browser tasks like form filling and data extraction from the command line?

Browser automation via the agent-browser CLI lets you open URLs, snapshot DOM elements into compact refs, fill forms, click controls, and extract data using simple Bash commands.

Can I run multiple isolated browser sessions with independent cookies and storage for testing?

Yes, browser automation supports multi-session execution with independent cookies, storage, and tabs, allowing you to run isolated web testing or save and restore session state for CI/CD pipelines.

What is a DOM snapshot ref and how does it help with web testing automation?

A DOM snapshot captures page elements and assigns them compact refs like @e1, enabling deterministic click and fill actions during web testing without relying on brittle CSS selectors.

Do I need a specific environment to run CLI-based browser automation and screenshots?

You need a Bash environment with the agent-browser CLI installed to execute browser automation commands, take screenshots, and run optional scripts or references.

What's the best way to capture website content and screenshots for documentation?

Use CLI browser automation to open target URLs, navigate pages, and capture screenshots directly to files like ./page.png, providing reproducible site content for documentation.

Are there limitations to using snapshot-based refs for form automation across multiple pages?

Snapshot-based refs are deterministic for the current page state but may change after page reloads or dynamic updates, requiring a new snapshot before continuing form automation on updated content.