ui-test

UI Test is an AI-powered adversarial UI testing skill designed to evaluate web applications in a real browser environment. Rather than simply verifying expected behavior, it focuses on identifying defects, regressions, usability issues, and edge-case failures. The skill can analyze code changes through git diffs to test only modified functionality, or perform broader exploratory testing across an application. It supports both local environments and remote testing workflows, helping teams validate UI quality before deployment.

The skill provides three primary testing workflows: diff-driven testing, exploratory testing, and parallelized testing across multiple browser instances. Testing is coordinated through a structured planning process that includes functional analysis, adversarial analysis, and coverage-gap analysis. It evaluates core user flows, error handling, edge cases, accessibility concerns, keyboard navigation, responsive layouts, console errors, and visual consistency. The skill organizes tests into groups, delegates execution to sub-agents, and consolidates results into a unified report. It emphasizes comprehensive bug discovery rather than early success confirmation.

This skill is intended for developers, QA engineers, accessibility auditors, and teams reviewing UI changes in pull requests. Common use cases include validating recent UI modifications, performing exploratory quality assurance, auditing accessibility compliance, testing responsive behavior across devices, and scaling browser-based test execution through parallel workflows. It is particularly useful when teams need systematic coverage of changed functionality while maintaining a focus on uncovering unexpected issues and regressions.