gstack

Automate QA sessions against web pages with a persistent headless browser.

Updated Apr 5, 2026
One-click install
npx skills add https://github.com/rahmanayon/beehive --skill gstack-rahmanayon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gstack
Source: https://github.com/rahmanayon/beehive/tree/main/.claude/skills/gstack
Command: npx skills add https://github.com/rahmanayon/beehive --skill gstack-rahmanayon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

gstack provides a fast headless browser-based QA environment that lets teams drive pages, interact with elements, capture bug evidence, and verify changes with diffed results.

Core Features & Use Cases

  • Navigate a page with commands like goto, back, forward, and reload
  • Read content via text, html, links, forms; snapshot with diffs and overlays
  • Interact with pages using click, fill, hover, type, and scroll
  • Generate annotated screenshots and diffs to verify UI changes across devices
  • Run end-to-end QA on staging sites and capture bugs with reproducible steps

Quick Start

Start by invoking /browse to launch the QA browser, load a page, and then use text, screenshot, and snapshot -D to verify changes.

Frequently Asked Questions about gstack

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

FAQPage Schema
How do I automate browser testing and capture screenshots for QA?

Automate browser testing by launching a persistent headless browser to navigate web pages, interact with elements, and generate annotated screenshots to capture bug evidence and verify UI changes.

What is visual regression testing and how do snapshot diffs work?

Visual regression testing verifies UI changes by comparing ref-based snapshots. It uses diffs and overlays to highlight visual differences between current and reference page states deterministically.

How do I run end-to-end QA sessions on a staging site?

Run end-to-end QA on staging sites by driving pages with commands like goto, back, forward, and reload, then interacting via click, fill, hover, type, and scroll to reproduce bugs.

Can I extract text and forms from web pages using a headless browser?

Yes, you can read content via text, html, links, and forms commands within the headless browser environment to extract deterministic data from loaded web pages.

Do I need any external dependencies to start headless browser automation?

No external dependencies are required. The environment relies on a Bun-based server and a ref-based snapshot system to deliver deterministic results directly.

What's the best way to verify UI changes across different devices?

Verify UI changes across devices by generating annotated screenshots and diffs using the snapshot command, which overlays visual differences to ensure consistent rendering.