browse

Automate QA checks on web applications with a persistent Playwright-based headless browser.

2|Updated Mar 16, 2026

Search Tags:#automation #playwright #headless #browse #dogfooding #qa-testing

Ready to run

npx skills add https://github.com/OmerFarukOruc/symphony-orchestrator --skill browse-omerfarukoruc

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: browse
Source: https://github.com/OmerFarukOruc/symphony-orchestrator/tree/main/.claude/skills/gstack/browse
Command: npx skills add https://github.com/OmerFarukOruc/symphony-orchestrator --skill browse-omerfarukoruc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Navigating and testing web interfaces is laborious without a repeatable, isolated environment. This Skill provides a persistent headless browser that AI agents can control to automate QA checks, reproduce bugs, and dogfood user flows locally.

Core Features & Use Cases

Deterministic browser actions: navigate to URLs, interact with elements, fill forms, handle dialogs, and verify page states in a sandboxed environment.
Evidence-first testing: diff before/after actions, capture annotated screenshots, test responsive layouts, and log browser activity for bug reports.
Use cases: verify a deployment by simulating a user journey, reproduce issues from issue trackers, or script end-to-end checks for regression testing.

Quick Start

Use the gstack browse skill to open a headless browser, navigate to a URL, and generate an evidence package (screenshots and state) for a bug report.

Frequently Asked Questions about browse

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

FAQPage Schema

How do I automate headless browser QA testing for web applications?▼

Automate headless browser QA testing by using a Playwright-based persistent browser to execute deterministic commands, interact with elements, and verify page states in a local sandbox. This allows AI agents to simulate user journeys and verify deployments automatically.

Can I capture screenshots and logs for bug reproduction during automated testing?▼

Yes, you can capture screenshots and logs for bug reproduction. The testing process supports evidence-first outputs, allowing you to diff before and after actions, generate annotated screenshots, and log browser activity to build comprehensive bug reports.

How do I reproduce user flow issues in an isolated environment?▼

Reproduce user flow issues by running automated end-to-end checks in a sandboxed headless browser with per-issue isolation. This ensures deterministic browser actions are tested without external interference, accurately replicating the reported bug.

Does Playwright support automated form filling and dialog handling for regression testing?▼

Playwright supports automated form filling and dialog handling for regression testing through deterministic browser actions. You can script end-to-end user flows, navigate to URLs, and interact with page elements to verify application stability.

What is the best way to verify a web deployment locally?▼

The best way to verify a web deployment locally is to script an automated user journey in a persistent headless browser. This simulates real user interactions, checks responsive layouts, and validates page states to ensure the deployment functions correctly.

When should I not use a headless browser for QA checks?▼

You should avoid using a headless browser for QA checks when you need to test visual rendering specific to a physical display, or when your application requires non-web protocols, as this approach is strictly designed for web interface automation and local sandbox dogfooding.