概要
The UX Evaluator skill provides a comprehensive framework for auditing digital products directly within Claude Code. It features three distinct modes: 'ux-eval' for lifecycle-based quality assessments, 'dogfood' for multi-layered production readiness checks (User, Tech, and Infra), and 'mcp-eval' for analyzing Model Context Protocol tool-to-widget chains. By integrating with Playwright for browser automation and Linear for issue tracking, it automates the discovery of friction points, technical bugs, and architectural inconsistencies, delivering structured markdown reports and actionable task lists that bridge the gap between design and implementation.
