gemstack

Automate browser-driven QA testing with a persistent headless browser.

1|3|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/ekrishgupta/gemstack --skill gemstack
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gemstack
Source: https://github.com/ekrishgupta/gemstack/tree/main
Command: npx skills add https://github.com/ekrishgupta/gemstack --skill gemstack

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, diff.

What problem does it solve?

Gemstack provides a fast, persistent headless browser and a workflow of skills to streamline QA testing, product dogfooding, and release readiness. It eliminates the friction of setting up and maintaining a browser-driven QA loop across local development and staging URLs.

Core Features & Use Cases

  • Automated browser-driven QA: goto, text, click, fill, and form interactions with a persistent session.
  • Diffed verification and annotated screenshots: compare before/after states and produce evidence for bug reports.
  • Dogfooding and design support: run /office-hours, /plan-design-review, /design-consultation to align product decisions with real usage.

Quick Start

To start, run the gemstack CLI and invoke /office-hours to frame your problem.

Frequently Asked Questions about gemstack

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

FAQPage Schema
How do I automate headless browser QA testing for staging and local development URLs?

Headless browser QA testing is automated by using a persistent Playwright-backed daemon to execute navigation, text extraction, click, fill, and form interactions. This enables repeatable automated browser-driven QA across local development and staging environments with sub-second command latency.

Can I capture and compare screenshots during e2e testing to verify before and after states?

Yes, you can capture annotated screenshots and perform diffed verification during e2e testing. The workflow allows you to compare before and after states directly, producing visual evidence for bug reports and release readiness.

How does a persistent headless browser improve QA automation latency compared to standard Playwright?

A persistent headless browser maintains a continuous browser daemon instead of launching new instances per test. This architecture provides sub-second command latency for QA automation and includes automatic idle shutdown to conserve resources.

Do I need to install Playwright and diff dependencies to run browser-driven QA workflows?

Yes, Playwright and diff are required dependencies to run browser-driven QA workflows. The Skill relies on a Bun-based binary powered by Playwright to manage the persistent browser daemon and perform diffed verification of content states.

What is the best way to start planning product dogfooding and design reviews for release readiness?

The best way to start planning product dogfooding and design reviews is to invoke the /office-hours command. This frames your problem and aligns product decisions with real usage through structured design consultation and review workflows.

Are there limitations when using headless browser automation for interaction verification across QA pipelines?

Limitations of headless browser automation include reliance on a persistent daemon that requires active management, though it features automatic idle shutdown. It is specifically designed for local development, staging, and QA pipelines rather than large-scale production scraping.