5-dim-review

Review completed designs across five dimensions with evidence-backed scores and prioritized findings.

Updated Aug 13, 2026
One-click install
npx skills add https://github.com/Mjlike123/ui-prototype-workbench --skill 5-dim-review-mjlike123
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: 5-dim-review
Source: https://github.com/Mjlike123/ui-prototype-workbench/tree/main/.agents/skills/monkren-design/skills/04-review/5-dim-review
Command: npx skills add https://github.com/Mjlike123/ui-prototype-workbench --skill 5-dim-review-mjlike123

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Design reviews often lack consistent criteria and concrete evidence, leading to subjective feedback. This Skill scores completed designs across five dimensions—philosophy consistency, visual hierarchy, detail execution, functionality, and innovation—with file, line, or element-level evidence for every score. ## Core Features & Use Cases - Five-Dimension Scoring: Rates designs 1-10 on philosophy consistency, visual hierarchy, detail execution, functionality, and innovation, with each score backed by observable evidence. - Scoped Review Depths: Supports quick reviews for single components and deep reviews for single-page or multi-page projects. - Optional Specialized Checks: Loads targeted modules on demand for AI-slop detection, accessibility auditing, hierarchy rhythm, and interaction state coverage. - Use Case: After building a prototype page, run a deep review to receive P0-P3 prioritized findings with locations, impact, evidence, and suggestions, plus a final five-dimension score summary. ## Quick Start Ask the AI to run a five-dimension design review on the completed page and report prioritized findings with evidence-backed scores.

Frequently Asked Questions about 5-dim-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a structured design review on a completed page?

Run a five-dimension review that scores philosophy consistency, visual hierarchy, detail execution, functionality, and innovation from 1 to 10. Each score must cite evidence such as a file, line number, element, or observable behavior, and findings are prioritized as P0 through P3.

What is the difference between a quick review and a deep design review?

A quick review targets a single component, while a deep review covers single-page or multi-page projects. The review depth is chosen during scope confirmation based on the object, medium, and key user goals.

Does the design review modify my files?

No, the review is read-only by default and only produces a report. File modifications happen only when the user explicitly requests fixes, which then hands off to a separate design improvement workflow.

Can the review check accessibility and interaction states?

Yes, specialized modules for accessibility auditing, interaction states, AI-slop detection, and hierarchy rhythm can be loaded on demand. They are run selectively based on task needs rather than unconditionally.

What happens when visual rendering tools are unavailable during review?

When rendering tools are unavailable, visual conclusions are explicitly limited to static code inference. The review states assumptions when information is insufficient instead of fabricating context.