Automates browser-based QA with persistent headless Chromium sessions and evidence outputs.
npx skills add https://github.com/MFS-code/gstack --skill gstack-mfs-code
Please help me install this Agent Skill. Skill: gstack Source: https://github.com/MFS-code/gstack/tree/main Command: npx skills add https://github.com/MFS-code/gstack --skill gstack-mfs-code
gstack solves the problem of slow, fragile QA and site dogfooding by giving agents a persistent headless Chromium browser to navigate, interact, verify UI state, and capture evidence like diffs and screenshots.
Use the gstack skill to QA a page by running your test flow on a real URL and capturing screenshots and snapshot diffs as evidence.
High-intent search queries and answers about installing and using this skill.
Automated UI verification captures screenshot evidence by navigating pages in a headless browser, interacting with elements, and saving annotated screenshots and snapshot diffs. This validates user flows and deployments while providing persistent session state for reliable testing.
A persistent headless browser maintains session state by keeping cookies, tabs, and session data active across sequential commands. This enables continuous multi-step QA testing flows without restarting login sessions or losing context between automated interactions.
Testing responsive layouts across viewports involves running automated browser workflows that resize the headless browser window and capture visual evidence. This process validates UI rendering and behavior across multiple screen dimensions using snapshot diffs and annotated screenshots.
Snapshot diffing uses ref-based selectors from accessibility snapshots to avoid brittle CSS guessing during element interaction. This method provides reliable element targeting by referencing structural page data rather than relying on volatile styling selectors that break during UI changes.
You can inspect console logs and network requests during automated browser workflows to verify application behavior and capture bug evidence. This inspection runs alongside navigation and element interaction within the persistent headless browser session.
Validating production deployments without manual site dogfooding is best achieved through automated headless browser workflows. These workflows navigate real URLs, interact with page elements, and generate unified diffs and annotated screenshots as persistent evidence of production behavior.
kentojeOpen an authenticated headless browser session for Aircall staging localhost testing.
JaykaijackAutomate headless Chromium browsing and UI verification with screenshots and diffs.
watzonAutomate browser tasks via headless CLI with ref-based element targeting.
cheng80Automate headless browser QA interactions and capture annotated screenshots.
