Design Is — Dieter Rams Design Audit
Audits a UI against Dieter Rams' ten principles with cited evidence, then issues a NEW / REFINE / REDESIGN verdict plus a ready-to-run `/make-plan` handoff.
Design & UIIntermediate★ 90,139⑂ 7,846AI score 8/10Last updated: Aug 9, 2026
What it does
Runs a structured design audit against Dieter Rams' ten "Good design is…" principles — innovative, useful, aesthetic, understandable, unobtrusive, honest, long-lasting, thorough, environmentally friendly, and as little design as possible.
- Phase 0 – Scope lock: records target surface, primary user, primary task and constraints in
00-scope.md - Phase 1 – Evidence fan-out: parallel subagents gather structural counts, spacing/type scales, contrast ratios, user-facing copy, JS weight, and accessibility facts — every finding must cite
file:lineor a screenshot region - Phase 2 – Scorecard: orchestrator (not subagents) scores each principle 0–3 using fixed anchors; ties break downward, worst instance wins over average, max 30
- Phase 3 – Verdict: ≥20 with no zeros → REFINE; <20 or a zero on a load-bearing principle → REDESIGN; nothing built yet → NEW
- Phase 4 – Handoff: emits one fully filled, self-contained
/make-planprompt
Everything lands in DESIGN-IS-<YYYY-MM-DD>/ as five markdown artifacts.
Who it's for
- Product designers and frontend engineers who want evidence-based critique instead of taste arguments
- Leads and PMs who must justify "refine" vs "start over" to stakeholders
- Solo builders checking for missing empty/loading/error/focus states and dark patterns
Examples
- "Design review of our settings page in
src/features/settings" → subagents count interactive elements and states, orchestrator scores 22/30 and returns a REFINE handoff with a regression checklist for the 3-scoring principles - "Is this checkout UI honest?" → the Copy & Honesty subagent lists every string with file:line, flags fake scarcity and hidden fees, principle #6 scores 0 → REDESIGN
- "No UI exists yet, I want a new onboarding flow" → Phases 1–2 are skipped, verdict = NEW, and a NEW DESIGN
/make-plantemplate is filled in with IA, token and state deliverables
· · · Install guide · · ·
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on Claude Skill Mart. Copy the plugin/skills/design-is folder from the GitHub repo thedotmack/claude-mem into my ~/.claude/skills/design-is/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/thedotmack/claude-mem.git /tmp/claude-mem && mkdir -p ~/.claude/skills && cp -r /tmp/claude-mem/plugin/skills/design-is ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal and clone the repo to a temp folder:
git clone https://github.com/thedotmack/claude-mem.git /tmp/claude-mem - Create the skills folder if needed:
mkdir -p ~/.claude/skills - Copy just this skill:
cp -r /tmp/claude-mem/plugin/skills/design-is ~/.claude/skills/ - Verify that
~/.claude/skills/design-is/SKILL.mdexists. - Restart Claude Code, then ask something like "audit this design" or "check this UI against Rams".
- (Recommended) Also install the repo's
/make-plancommand andagent-browserskill so the handoff and screenshot steps work; without them you can still use the audit artifacts.
View source on GitHub ↗License: Apache-2.0