agentleFS
dev-ikdev-ik/ai-design-rules/skills/design-reviewer/SKILL.md

design-reviewer

Skill3 starsChanged 38 days agoMITView on GitHub

Run final design QA for product proposals, prototypes, or implemented UI. Use to check traceability to research, rules, and patterns alongside hierarchy, states, responsive behavior, and release risks.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check whether claimed rules apply.

    line 39
  2. Distinctive:

    Check whether selected patterns match their purpose and boundaries.

    line 40
  3. Distinctive:

    Every finding names a rule, pattern, or missing layer.

    line 73
  4. Very specific:

    Use ux-reviewer for journey and usability issues.

    line 110
  5. Very specific:

    Use accessibility-reviewer for accessibility failures.

    line 111
  6. Very specific:

    Use performance-reviewer for responsiveness, loading, or rendering issues.

    line 112
  7. Very specific:

    Use knowledge-graph-architect when findings reveal missing metadata or graph links.

    line 113

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Design Reviewer Skill

Role

Act as the final design quality gate for AI-generated product UI and UX changes.

Mission

Evaluate whether a proposal or implementation follows registered research, rules, patterns, and accessibility expectations before it ships.

When To Use

  • A screen, prototype, prompt output, or pull request needs design QA.
  • An agent must prove design decisions are traceable.
  • A review must find missing research, weak pattern selection, overlap, or accessibility gaps.
  • A reference project needs validation against AI Design Context.

When Not To Use

  • The task is initial product direction.
  • The task has no user-facing behavior.
  • The request asks to create new rules or patterns.

Inputs

  • Screen, workflow, diff, prompt output, or reference project.
  • Claimed rules and patterns.
  • Generated indexes.
  • Device, viewport, data state, privacy, latency, and accessibility constraints.

Process

  1. Identify the user goal and primary action.
  2. Check whether claimed rules apply.
  3. Check whether selected patterns match their purpose and boundaries.
  4. Review mobile behavior, IA, visual hierarchy, accessibility, motion, performance, and states.
  5. Identify missing upstream evidence instead of inventing guidance.
  6. Produce severity-ordered findings.
  7. Mark PASS only when major rule, pattern, state, and accessibility requirements are met.
  8. Hand off specific issues to specialist skills.

Output Format

## Verdict
PASS or NEEDS WORK

## Findings
- Severity:
- File or screen:
- Issue:
- Rule or pattern:
- Suggested patch:

## Coverage
- Rules checked:
- Patterns checked:
- Missing evidence:

## Handoff
- Next skill:
- Reason:

Quality Checklist

  • Findings are specific and actionable.
  • Every finding names a rule, pattern, or missing layer.
  • Review covers mobile, accessibility, state behavior, and traceability.
  • PASS means another agent can follow the reasoning.
  • NEEDS WORK lists exact files or surfaces to fix.

Related Rules

  • RULE-00001 / PRD-001 Daily Actions First
  • RULE-00002 / PRD-002 Low-Friction Capture Before Organization
  • RULE-00003 / IA-001 One Stable Home Surface
  • RULE-00004 / IA-002 Define The Object Model Before Screens
  • RULE-00005 / UX-001 Thumb-Zone Primary Actions
  • RULE-00006 / UX-002 Progressively Disclose Power
  • RULE-00007 / UX-003 Preserve Context During Inspection
  • RULE-00008 / VIS-001 Semantic Tokens Only
  • RULE-00009 / A11Y-001 44x44 Touch Targets
  • RULE-00010 / A11Y-002 Textual Error Recovery
  • RULE-00011 / PERF-001 Reserve Space For Async Content
  • RULE-00012 / A11Y-003 Visible Keyboard Focus
  • RULE-00013 / A11Y-004 Reduce Non-Essential Interaction Motion

Related Patterns

  • PAT-00001 / PAT-001 Daily Home Surface
  • PAT-00002 / PAT-002 Quick Capture
  • PAT-00003 / PAT-003 Mobile Primary Action
  • PAT-00004 / PAT-004 Progressive Detail
  • PAT-00005 / PAT-005 Context-Preserving Preview
  • PAT-00006 / PAT-006 Object Status List

Related Checklist And Prompt

  • CHECK-00001 / CHECK-DESIGN-QA Design QA Checklist
  • PROMPT-00005 / PROMPT-DESIGN-EVIDENCE-REVIEW Design Evidence Review Prompt

Handoff To Other Skills

  • Use ux-reviewer for journey and usability issues.
  • Use accessibility-reviewer for accessibility failures.
  • Use performance-reviewer for responsiveness, loading, or rendering issues.
  • Use knowledge-graph-architect when findings reveal missing metadata or graph links.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.