agent-browser

Automate browser tasks via CLI with ref-based element interactions.

1|Updated Apr 13, 2026
One-click install
npx skills add https://github.com/weorbitant/compound-engineering-feat-python-plugin --skill agent-browser-weorbitant
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/weorbitant/compound-engineering-feat-python-plugin/tree/main/skills/agent-browser
Command: npx skills add https://github.com/weorbitant/compound-engineering-feat-python-plugin --skill agent-browser-weorbitant

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates browser-based tasks from the command line, enabling AI agents and developers to interact with web pages, fill forms, take screenshots, or scrape data without a GUI.

Core Features & Use Cases

  • CLI-driven browser automation with ref-based element interactions (@e1, @e2) and snapshot-based element discovery.
  • Navigate, fill, click, and extract data across dynamic web apps for automation, testing, QA, and data collection.
  • Use cases include automated form submission, web data extraction, and end-to-end interaction flows in single-page and multi-page sites.

Quick Start

Open a URL with agent-browser and begin interacting with page elements using 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 browser interactions from the command line?

Browser automation from the command line is achieved by using the agent-browser CLI to navigate web pages, fill forms, click elements, and capture data without a GUI. It applies ref-based element selection and snapshots to drive interactions across dynamic web apps.

Can I run automated web scraping and form submission on single-page apps?

Automated web scraping and form submission work on single-page apps by leveraging snapshot-based element discovery and ref-based interactions. The CLI navigates dynamic content, fills forms, and extracts data across multi-step flows within single-page and multi-page sites.

What is ref-based element selection in CLI browser testing?

Ref-based element selection in CLI browser testing identifies page elements using references like @e1 and @e2. This mechanism targets specific elements discovered via page snapshots, enabling precise scripted interactions for rapid QA and automated testing.

Do I need agent-browser installed globally to automate web tasks?

Yes, you need the agent-browser tool installed globally to automate web tasks. This prerequisite environment setup allows the CLI to execute browser-based interactions, take screenshots, and perform data extraction across web pages.

What's the best way to capture screenshots during multi-step web automation?

The best way to capture screenshots during multi-step web automation is using the agent-browser CLI. It supports screen capture across scripted interaction flows, allowing you to snapshot page states during rapid QA and end-to-end testing.

Why use CLI browser automation instead of GUI-based testing tools?

CLI browser automation enables AI agents and developers to execute web interactions, form submissions, and data extraction without a GUI. It supports rapid QA and automated testing across single-page apps, making it suitable for headless environments and scripted workflows.