skills.rest
compojoom
compojoomCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

agent-browser

Automate browser interactions via a Bash CLI using ref selectors.

Updated Aug 23, 2026
Search Tags:#automation#CLI#web automation#browser#agent-browser#ref selectors
One-click install
npx skills add https://github.com/compojoom/safe-engineering-plugin --skill agent-browser-compojoom
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/compojoom/safe-engineering-plugin/tree/main/plugins/safe-engineering/skills/agent-browser
Command: npx skills add https://github.com/compojoom/safe-engineering-plugin --skill agent-browser-compojoom

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Browser automation for AI agents using a CLI to open pages, interact with elements via ref-based selectors, and capture results.

Core Features & Use Cases

  • CLI-based browser automation using ref selectors (@eN) to target interactive elements.
  • Navigate, click, fill, type, take screenshots, and extract data across web pages.
  • Use cases include automating login flows, form submissions, data scraping, and quick QA checks.

Quick Start

Open a page: agent-browser open https://example.com Snapshot the page to identify refs: agent-browser snapshot -i Interact with elements: agent-browser click @e1; agent-browser fill @e1 "sample"; agent-browser type @e1 "text" Save a screenshot: agent-browser screenshot output.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 web tasks with CLI browser control?▼

You can automate web tasks by using a Bash-based CLI to open pages, snapshot interactive elements, and execute actions like click, fill, and type via ref selectors. It supports navigation, screenshots, and data extraction for quick web automation.

What are ref selectors and how do they work for web automation?▼

Ref selectors (@eN) are semantic locators used to target interactive elements on a web page. You snapshot a page to identify these refs, then use them in CLI commands to click, fill, or type into specific elements for automated interactions.

Can I use this for automating login flows and form submissions?▼

Yes, automating login flows and form submissions is a core use case. The CLI provides actions to open pages, fill form fields, type text, and click buttons, enabling you to script complete automated interaction sequences.

How do I extract data or capture screenshots from a web page?▼

You can extract data by taking a snapshot with JSON output, or capture visual results using the screenshot command to save a PNG. The snapshot command identifies interactive elements, while screenshots save the current page state.

Does this browser automation tool require any external dependencies?▼

No external dependencies are required. The tool operates independently through a Bash-based CLI, providing built-in session management and core browser actions without needing additional packages or frameworks installed.

What are the limitations of using a CLI for browser automation?▼

It is optimized for simple and single-page websites. While it handles navigation, form filling, and data extraction, complex multi-tab workflows or highly dynamic single-page applications may exceed its snapshot and ref selector targeting capabilities.

Related Skills

Explore More Software Engineering Skills→
ATaC-teamATaC-team
291

atac

Build and execute AI agent workflows with a declarative trajectory CLI.

Community
Advanced
zhu637882-stackzhu637882-stack

scrape

Extract structured data from web pages into JSON payloads.

Community
Intermediate
k97k97
4

agent-browser

Automate browser navigation, form submission, and data extraction for web testing.

Community
Advanced
tharunramagiritharunramagiri

agent-browser

Automate web navigation, element interaction, and data extraction via a headless browser.

Community
Advanced