gstack

Control a persistent headless Chromium daemon via a local CLI for browser testing.

Updated Mar 28, 2026
One-click install
npx skills add https://github.com/24601/gastack-pre-fork --skill gstack-24601
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gstack
Source: https://github.com/24601/gastack-pre-fork/tree/main
Command: npx skills add https://github.com/24601/gastack-pre-fork --skill gstack-24601

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, puppeteer-core, and includes scripts (resource) components.

What problem does it solve?

Many browser-driven QA and dogfooding workflows are slow, flaky, and lose session state because each tool launches a fresh browser. gstack provides a persistent headless Chromium daemon with ref-based element selection so agents and CLI workflows run sub-second commands, preserve cookies/tabs, and produce reliable evidence (annotated screenshots, diffs, logs) for testing and verification.

Core Features & Use Cases

  • Persistent daemon: long-lived Chromium process that avoids repeated cold starts and preserves cookies, tabs, and localStorage across commands.
  • Ref-based interaction: snapshot assigns @e/@c refs to elements so agents can click/fill without brittle CSS selectors.
  • Verification tools: snapshot diffs (-D), annotated screenshots (-a -o), responsive screenshots, perf metrics, console/network/dialog buffers, and multi-step chain execution.
  • Auth & handoff: import real browser cookies, headed-mode (connect) for live handoff, and handoff/resume to let a human solve CAPTCHAs or MFA then continue automated flows.
  • Use cases: QA regression testing, deployment verification, form flows with authentication, visual diffs for PRs, and automated multi-step user journeys.

Quick Start

Launch the gstack browse binary and use its $B commands (goto, snapshot, click, screenshot) to open a page, capture interactive refs, and save an annotated screenshot.

Frequently Asked Questions about gstack

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

FAQPage Schema
How do I keep browser sessions persistent for Playwright QA automation?

You can keep browser sessions persistent for Playwright QA automation by running a long-lived headless Chromium daemon that preserves cookies, tabs, and localStorage across sub-second CLI commands, eliminating cold starts.

How do I automate browser testing without brittle CSS selectors?

Automate browser testing without brittle CSS selectors by using a snapshot command that assigns interactive refs to page elements, allowing agents to click or fill directly via stable references.

Can I import real browser cookies into a headless Chromium daemon?

Yes, you can import real browser cookies into a headless Chromium daemon to handle QA flows and deployment verification requiring authenticated multi-step testing across staging and production-like environments.

What is the best way to handle CAPTCHAs during automated browser testing?

The best way to handle CAPTCHAs during automated browser testing is using headed-mode handoff to let a human solve the challenge, then resume automated multi-step user journeys in the persistent daemon.

How do I capture visual regression evidence with annotated screenshots?

Capture visual regression evidence with annotated screenshots and snapshot diffs by running verification commands that output responsive screenshots, perf metrics, and console buffers from the persistent headless browser.

Does gstack work with Puppeteer for headless browser automation?

Yes, gstack works with Puppeteer for headless browser automation, building its persistent Chromium daemon and ref-based element selection on both Playwright and puppeteer-core dependencies.