skills.rest
Firzus
FirzusCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

agent-browser

Automate live browser interactions and capture screenshots via CDP.

1|Updated May 12, 2026
Search Tags:#browser automation#web testing#ui review#cdp#chrome#screenshot#accessibility tree
One-click install
npx skills add https://github.com/Firzus/agent-skills --skill agent-browser-firzus
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/Firzus/agent-skills/tree/main/skills/engineering/agent-browser
Command: npx skills add https://github.com/Firzus/agent-skills --skill agent-browser-firzus

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill lets you interact with live websites and browser-based apps from the command line without relying on brittle HTML parsing or manual clicking.

Core Features & Use Cases

  • Browser automation: Open pages, navigate, click, fill forms, type, select, and capture screenshots.
  • Reliable interaction loop: Use accessibility-tree snapshots with stable element refs, then re-snapshot after page changes to avoid stale targets.
  • Testing and review workflows: Inspect UI behavior, measure layout and performance, check console errors, and compare screenshots for regressions.
  • CDP attachment: Connect to existing Chromium-based apps such as Chrome, Tauri/WebView2, and Electron through a remote debugging endpoint.

Quick Start

Ask the assistant to open a target website in agent-browser, snapshot the page, and perform the next interaction using the element references it returns.

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 interaction from the CLI without brittle HTML parsing?▼

Browser automation from the CLI uses accessibility-tree snapshots with stable element references to interact with live pages, navigating and filling forms without relying on brittle HTML parsing or manual clicks.

How does accessibility tree snapshotting handle stale element references during web testing?▼

Accessibility tree snapshotting handles stale references by capturing stable element identifiers, then re-snapshotting the page after each interaction to avoid stale targets during web testing workflows.

Can I connect to existing Chromium apps like Electron or Tauri for UI inspection?▼

Yes, you can connect to existing Chromium apps like Electron, Tauri, and WebView2 for UI inspection by attaching to their remote debugging endpoints via the Chrome DevTools Protocol (CDP).

What is the best way to capture UI evidence and check for layout regressions?▼

The best way to capture UI evidence and check for layout regressions is using agent-driven browser control to take screenshots, inspect console errors, and compare visual output automatically.

Why do I need persistent browser sessions for agent-driven browser control?▼

Persistent browser sessions are required for agent-driven browser control to maintain page state across multi-step interactions, ensuring reliable navigation, form filling, and sequential UI testing.

Does browser automation work with Chrome DevTools Protocol for performance checks?▼

Yes, browser automation works with the Chrome DevTools Protocol (CDP) for performance checks, allowing direct attachment to Chromium instances to measure layout metrics and inspect runtime behavior.

Related Skills

Explore More Software Engineering Skills→
yuexpsyuexps
3

browser_cdp

Connect to Chrome browsers via CDP and scan debugging ports.

Community
Intermediate
tech-with-sethtech-with-seth
4

playwright-cli

Automate browser UI tests using Playwright CLI commands.

Official
Advanced
johnlindquistjohnlindquist
16

chrome-devtools

Automate browser control and debugging via the Chrome DevTools Protocol.

Community
Advanced
AradotsoAradotso
8

adspower-antidetect-browser

Manage and automate AdsPower antidetect browser profiles via local HTTP API.

Official
Advanced