agent-browser

Automate browser interactions to open websites, fill forms, and extract data.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/indigoai-us/hq --skill agent-browser-indigoai-us
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/indigoai-us/hq/tree/main/template/workers/public/qa-tester/skills/agent-browser
Command: npx skills add https://github.com/indigoai-us/hq --skill agent-browser-indigoai-us

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Automate browser interactions for AI agents to open websites, fill forms, navigate pages, click elements, and extract data, enabling hands-off web automation and testing.

Core Features & Use Cases

  • Snapshot-driven refs: capture compact element references for reliable automation without verbose selectors.
  • Multi-session isolation: run concurrent browser sessions with independent cookies, storage, and history.
  • Form automation & data extraction: fill forms, click controls, scrape text, and export results with minimal prompts.
  • Templates & workflows: reusable templates (e.g., authenticated sessions, capture workflows) to speed up common tasks.

Quick Start

Open a website with agent-browser, fill a form, and snapshot the page to verify results.

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 interactions like form filling and page navigation with AI agents?

Automate browser interactions by using terminal commands to open websites, fill forms, navigate pages, and click elements. AI agents capture snapshot-driven refs for reliable automation without verbose selectors, streamlining web testing and data extraction.

Can I run concurrent browser sessions with isolated cookies and storage for web testing?

Concurrent browser sessions support multi-session isolation, maintaining independent cookies, storage, and history. This allows simultaneous web testing and data collection tasks across isolated environments without state interference.

What is the best way to extract data from web pages without writing verbose CSS selectors?

Data extraction uses snapshot-driven refs to capture compact element references, replacing verbose selectors. This enables reliable scraping and form automation directly through terminal commands with minimal prompts.

Does agent-browser work with local and cloud browsers for QA tasks?

Agent-browser operates against both local and cloud browsers for QA tasks. It exposes terminal commands to open, snapshot, fill, and click, supporting state persistence and reusable templates for authenticated sessions across web apps.

How do I fill a form and verify the result using snapshot-driven browser automation?

Fill forms by executing the fill command against targeted element refs, then use the snapshot command to capture the page state. This verifies automation results quickly without manual inspection.

What are the limitations of using snapshot-driven refs for web automation?

Snapshot-driven refs rely on capturing current page elements, so dynamic content loading or frequent DOM changes may require repeated snapshots. State persistence and templates help maintain context but cannot handle highly volatile single-page applications seamlessly.