LogoSkills

critique-ux

Run a focused UX critique on a screen โ€” affordances, information density, and hierarchy โ€” and output a prioritised fix list.

ํ•ญ๋ชฉ๋‚ด์šฉ
Invoke/critique-ux
argument-hint [screen name, Figma URL, or image โ€” e.g., 'checkout step 3' or a screenshot]

/critique-ux โ€” ๋น ๋ฅธ ๊ธฐ๋Šฅ ์ค‘์‹ฌ UX ์ ๊ฒ€#

ํ•œ๋งˆ๋””๋กœ#

ํ™”๋ฉด ํ•˜๋‚˜์—์„œ "์ผ์„ ๋๋‚ผ ์ˆ˜ ์žˆ๋Š”๊ฐ€"์— ์ง‘์ค‘ํ•ด, ์–ดํฌ๋˜์Šคยท์ •๋ณด ๋ฐ€๋„ยท์‹œ๊ฐ ์œ„๊ณ„ ์„ธ ๊ฐ€์ง€๋งŒ ๋น ๋ฅด๊ฒŒ ์ ๊ฒ€ํ•˜๊ณ  ์šฐ์„ ์ˆœ์œ„ ์ˆ˜์ • ๋ชฉ๋ก์„ ๋ฝ‘์•„์ฃผ๋Š” ๊ฐ€๋ฒผ์šด ๋ฒ„์ „์˜ ํฌ๋ฆฌํ‹ฑ ๋ช…๋ น์–ด์ž…๋‹ˆ๋‹ค.

๋ˆ„๊ฐ€ยท์–ธ์ œ ์“ฐ๋‚˜์š”#

  • ๋น ๋ฅธ ํ”ผ๋“œ๋ฐฑ ๋ฃจํ”„๊ฐ€ ํ•„์š”ํ•˜๊ฑฐ๋‚˜, PM์ด ์ฃผ๋„ํ•˜๋Š” ๋ฆฌ๋ทฐ์—์„œ ์ „์ฒด ์‹œ๊ฐ ๊ฐ์‚ฌ ์—†์ด ๊ธฐ๋Šฅ์ ์ธ ์ ๊ฒ€๋งŒ ๋ฐ›๊ณ  ์‹ถ์„ ๋•Œ
  • ์ „์ฒด 7์ข… ํฌ๋ฆฌํ‹ฑ(/critique-screen)๊นŒ์ง€๋Š” ํ•„์š” ์—†๊ณ , ์ž‘์—…์„ ๋ง‰๋Š” ๋ฌธ์ œ๋งŒ ๋น ๋ฅด๊ฒŒ ์ฐพ๊ณ  ์‹ถ์„ ๋•Œ

๋ฌด์—‡์„ ํ•ด์ฃผ๋‚˜์š”#

  • ์‹œ๊ฐ ์œ„๊ณ„(์ฃผ์š” ์•ก์…˜์ด ๋ช…ํ™•ํ•œ์ง€), ์–ดํฌ๋˜์Šค, ์ •๋ณด ๋ฐ€๋„ ์„ธ ๊ฐ€์ง€ ๊ด€์ ๋งŒ ์ ๊ฒ€ํ•ฉ๋‹ˆ๋‹ค.
  • ๋ฌธ์ œ๋ฅผ P1/P2/P3๋กœ ์šฐ์„ ์ˆœ์œ„ ๋งค๊ฒจ ์ •๋ฆฌํ•ฉ๋‹ˆ๋‹ค.
  • ํ™”๋ฉด์—์„œ ๊ฐ€์žฅ ํฐ ๊ธฐ๋Šฅ์  ๋ฆฌ์Šคํฌ๋ฅผ ํ•œ ๋ฌธ์žฅ์œผ๋กœ ์š”์•ฝํ•ด ์ค๋‹ˆ๋‹ค.

โš™๏ธ ์ƒ์„ธ ์˜ต์…˜ยท์‹คํ–‰ ๋ช…์„ธ (๊ฐœ๋ฐœ์ž / AI ์—์ด์ „ํŠธ์šฉ)

Run a focused functional critique on a screen. This command covers the three dimensions most likely to block task completion: whether actions are discoverable, whether the right information is present, and whether priority is clear. Use this for quick feedback loops, PM-led reviews, or when you want functional critique without a full visual audit.

Steps#

  1. Visual hierarchy โ€” Identify entry point and whether the primary action reads clearly using critique-visual-hierarchy skill.
  2. Affordance โ€” Evaluate clickability signals, state visibility, CTA clarity, and action discoverability using critique-affordance skill.
  3. Information density โ€” Assess cognitive load, content priority, scanning patterns, and progressive disclosure using critique-information-density skill.
  4. Prioritise โ€” Collect every flagged issue and rank them:
    • P1 โ€” Critical: Blocks task completion or hides a required action; fix before shipping.
    • P2 โ€” Important: Slows the user or creates confusion; fix in current sprint.
    • P3 โ€” Polish: Minor friction or clarity improvement; address when capacity allows.

Output#

A prioritised fix list grouped by level. Each item includes:

  • Issue: what is wrong
  • Dimension: Hierarchy / Affordance / Density
  • Fix: the specific change required Conclude with one sentence summarising the biggest functional risk on the screen.