Can I skip specific testing perspectives?

Yes, the skill supports a --skip flag allowing you to bypass specific perspectives such as visual, functional, hig, edge, or ux if they are not required for your current check.

How does the team lead agent coordinate the testing?

The qa-lead agent creates a detailed task list, manages execution dependencies between testers, and aggregates all individual reports into a final summary with a composite score.

Where are the test results and screenshots stored?

All verification artifacts, including markdown reports and supporting screenshots, are automatically saved to the claudedocs/qa/ directory in your project root.

What platforms does the QA Agent Team support?

It supports Web (via Chrome and Playwright), Electron, iOS/Expo (via Simulator), and native macOS applications using platform-specific MCP servers.

What is the 'Quick' mode?

The --quick flag streamlines the process by deploying only three teammates (lead, visual, and functional) instead of the full six-agent squad for faster validation.

QA Agent Team

Name: QA Agent Team
Author: laststance

bylaststance

0•

보안 및 테스팅

Deploys a specialized team of autonomous AI agents to perform comprehensive end-to-end quality assurance across visual, functional, and UX dimensions.

The QA Agent Team skill orchestrates a multi-disciplinary squad of AI agents to validate software implementations from five distinct perspectives: Visual Integrity, Functional Correctness, Apple HIG Compliance, Edge Case robustness, and UX Sensibility. By leveraging platform-specific MCP tools like Playwright, Electron, and iOS simulators, it automates rigorous verification workflows, generates detailed markdown reports with screenshots, and provides a final composite quality score. This skill is ideal for post-implementation verification, ensuring that new features meet production-grade standards and maintain cross-platform consistency without manual overhead.

주요 기능

01Rigorous 4-Phase Impact Propagation Protocol for deep functional validation.

02Comprehensive reporting with aggregated scoring and automated quality gate verdicts.

030 GitHub stars

04Platform-aware verification for Web, Electron, iOS/Expo, and macOS applications.

05Automated screenshot capture and RALPH protocol visual integrity checks.

06Multi-agent orchestration with specialized roles for visual, functional, and UX testing.

사용 사례

01Validating cross-platform UI consistency and Apple HIG compliance for mobile or desktop apps.

02Performing a final quality gate check after a major feature implementation or workflow completion.

03Identifying edge-case bugs and UX friction points that manual testing might overlook.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add laststance/claude-code-marketplace qa-team

For use in Claude.ai and ChatGPT

Download Skill