skills.rest
tony30552001
tony30552001COMMUNITY
Explore all skills by this author →
GitHub

gstack

Automates headless browser QA with scriptable click, fill, and screenshot workflows.

Updated Dec 26, 2025
Search Tags:#automation#security#qa#playwright#browser#claude-code
One-click install
npx skills add https://github.com/tony30552001/Genpic-master --skill gstack-tony30552001
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gstack
Source: https://github.com/tony30552001/Genpic-master/tree/main/.agents/skills/gstack
Command: npx skills add https://github.com/tony30552001/Genpic-master --skill gstack-tony30552001

SYSTEM DOCUMENTATION & REQUIREMENTS

## What problem does it solve? Claude Code users often need reliable, repeatable browser testing and QA workflows. gstack provides a persistent, deterministic browser runtime and a structured skill contract to drive real browser sessions from Claude.

## Core Features & Use Cases

  • Deterministic, headless browser interactions (goto, click, fill, etc.) via a per-skill runtime.
  • Ref-based element selection using the accessibility tree for stable, framework-agnostic targeting.
  • Side Panel/Real Browser workflows for design reviews, QA, and live-site validation; supports pair-agent and tunnel isolation.
  • Security guards including token-based auth, prompt-injection defenses, and data privacy safeguards.

### Quick Start Invoke the /qa workflow on a staging URL to begin automated browser testing and capture results.

Frequently Asked Questions about gstack

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

FAQPage Schema
How do I automate browser-based QA and testing workflows in Claude Code?▼

Automate browser-based QA in Claude Code by invoking the /qa workflow on a staging URL, which drives a persistent, deterministic headless browser daemon to execute repeatable goto, click, fill, and screenshot interactions.

What is ref-based element selection for browser automation?▼

Ref-based element selection uses the accessibility tree to target elements, providing a stable, framework-agnostic locator system for deterministic browser interactions instead of relying on fragile CSS selectors.

Can I use Playwright to run live site validation and design reviews?▼

Yes, you can run live site validation and design reviews using side-panel workflows, which support pair-agent configurations and tunnel isolation to validate real browser sessions directly from Claude.

Does this browser testing runtime include prompt-injection defense?▼

Yes, the browser testing runtime implements layered prompt-injection defenses alongside a strict, token-scoped authentication model to protect data integrity and prevent malicious commands during automated testing.

How do I capture snapshot diffs during automated browser debugging?▼

Capture snapshot diffs during automated browser debugging by applying the deterministic command surface to execute repeatable interactions, allowing you to compare visual states across different testing stages.

What are the limitations of using a headless browser daemon for testing?▼

Using a headless browser daemon for testing requires managing a persistent runtime and may not fully capture complex side-panel real browser workflows or visual rendering nuances without explicit screenshot and snapshot diff implementations.

Related Skills

Explore More Software Engineering Skills→
him55710-sudohim55710-sudo

gstack

Run a persistent Playwright headless browser daemon for Claude Code workflows.

Community
Advanced
espensevespensev

ui-test-engineer

Triage visual regressions in multimodal UI test evidence to identify root causes.

Community
Intermediate
joneqianjoneqian

agent-browser

Control a headless browser using accessibility-tree snapshots and ref-based element targeting.

Community
Advanced
EhsaanArkEhsaanArk

browse

Automates headless browser QA with Playwright, capturing diffs and screenshots.

Community
Advanced