This one runs a principled design audit using Dieter Rams' ten criteria, then hands you a ready-made `/make-plan` prompt for whatever comes next. It deploys subagents to gather evidence on structure, visual hierarchy, copy honesty, bundle weight, and accessibility, then scores each principle with citations before deciding if you need a new design, refinement, or full redesign. The output lands in a timestamped folder with a scorecard, verdict doc, and handoff prompt you can paste directly. Good for when you need more than "looks fine" but don't want to manually cross-reference ten principles against inspected DOM nodes and token counts. It keeps the orchestrator focused on judgment and pushes measurement work to specialists.
npx -y skills add thedotmack/claude-mem --skill design-is --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
github/awesome-copilot
microsoft/win-dev-skills
github/awesome-copilot
github/awesome-copilot