Reviewing design work
You are judging existing design output — a PR diff, a rendered surface, a
screenshot, a Figma handoff, or a spec — not building a system and not consuming one.
This is an L1 dispatcher: it orients you, orders the lenses, and points at the skill
that carries each lens's rules. Run the lenses in the order below so objective rule
violations surface before subjective bar judgments, and cheap code reads run before
expensive browser-driving passes. Each station gets a pointer, not a re-teaching — load
the station's own skill to actually run it.
When to use
- Reviewing a PR that adds or changes UI code, styles, tokens, or theme.
- Critiquing a rendered surface or screenshot (every theme and mode).
- Reviewing a design spec or a Figma → code handoff before the build lands.
- Running the design-critic gate on a slice (see orchestrating-elite-agent-qa).
When NOT to use
- Building a design system (tokens, scales, palettes, naming) — load
designing-a-design-system, not this dispatcher.
- Consuming a design system (wiring an existing token set into a product) — load
consuming-a-design-system.
- Pure logic / backend review with no design surface.
Lens order
Run the lenses in this order within a review session. (In clud-bug's pipeline the
code and design lenses are independently gated passes — SPEC §2.2 — so this order
sequences a reviewer's work and findings, not the passes.)
Code + markup lens — web-interface-guidelines-review. The opinionated rule set
(WIG, Material 3, Radix) plus token-driven contrast/typography/spacing rules: APCA-preferred
contrast cross-checked with WCAG 2.2 AA, atomic typography class, token-not-raw-hex,
verb-noun labels, the focus-ring contract, the 24 CSS px interactive floor. Findings
cite rules, not vibes. REQUIRED BACKGROUND: web-interface-guidelines-review.
Rendered-surface lenses — on screenshots in every theme and mode. Three lenses judge the
render itself:
- design-system-consistency — token / scale / color-discipline drift visible on
the render, not just the source line.
- frontend-a11y — contrast ratios, focus visibility, tap targets, semantics, and
motion on the rendered surface, in every mode.
- visual-polish — alignment, optical centering, spacing rhythm, glyph/pattern
quality, state coverage, theme parity ("fine but not elite" counts).
REQUIRED BACKGROUND: design-system-consistency, frontend-a11y, visual-polish.
Opinion lens — designing-elite-ui. The concrete elite / Figma-grade bar
(one-axis color, gated contrast, subtraction, on-grid sizes, every mode). It runs
after the objective lenses because it is the subjective bar; the objective
violations should already be caught. REQUIRED BACKGROUND: designing-elite-ui.
House stance last — when the work is that house's. Apply the L2 stance the
work belongs to — thrill-design for Thrill's; it leans
on the bar where the bar leaves room and never overrides an objective finding.
Skip for anyone else's work.
Each lens judges against the L0 primitives it does not re-derive — the color / type /
spacing math: oklch-color-space, apca-contrast, wcag-contrast,
palette-relationships, chroma-harmonization, type-scale, line-height-grid,
token-naming-conventions, component-sizing-principles.
A finding justified by a named principle rather than by what was observed has two more
L0 primitives behind it: empirical-design-principles
for the falsifiable ones — Gestalt proximity (it carries the ratio a grouping claim must
clear, which the polish lens defers to), Fitts, Hick, Miller, von Restorff — and
usability-heuristics for Nielsen, Norman and
Shneiderman, where the gate is source and edition rather than prediction.
What fires when
| Situation |
What fires |
| PR touches UI code, no visual surface change (routing, state wiring, refactor) |
Code lens only; the rendered lenses stay silent. |
| PR changes a visual surface (component, layout, styles, theme) |
Code lens and rendered lenses, with designing-elite-ui, then any house stance, after the objective lenses; the rendered pass is browser-driven — screenshots in every mode, states exercised — per orchestrating-elite-agent-qa (REQUIRED BACKGROUND for orchestrating that gate). |
| Design spec / Figma handoff, no code yet |
web-interface-guidelines-review applies to the spec; designing-elite-ui sets the bar the build must hit; the rendered lenses defer to post-build review. |
| clud-bug-installed repo |
The 5 kind: design skills (four lenses and the house stance: design-system-consistency, frontend-a11y, visual-polish, designing-elite-ui, thrill-design) run as the separate design pass (SPEC §4.8), never inline with the code review, and only when the repo opts in: the design pass is listed in review.passes in .clud-bug.json, at least one design skill applies to the diff, and the trigger is a pull request. Installation alone never fires the pass, and design findings are advisory unless the repo marks that pass blocking. |
System-specific layers
These lenses are generic; an L2 stance composes on top rather than replacing them.
thrill-design is one house's (lens 4). udts-review is one
system's (incubating): a parity marker pointing at tokenomics, not loadable guidance yet.
Cross-references
- L0 primitives: oklch-color-space, apca-contrast, wcag-contrast,
palette-relationships, chroma-harmonization, type-scale, line-height-grid,
token-naming-conventions, component-sizing-principles,
empirical-design-principles, usability-heuristics, motion-continuity.
- L0 critic lenses: web-interface-guidelines-review, design-system-consistency,
frontend-a11y, visual-polish, designing-elite-ui; and orchestrating-elite-agent-qa
(rendered-pass gate).
- L1 siblings: designing-a-design-system, consuming-a-design-system.
- L2 stances: udts-review (incubating), thrill-design.
1---2version: "1.1.0"3digest: "ee642a5b1da4" # is your copy current? github.com/thrillmade/agent-skills/blob/main/docs/skill-versions.json4origin: "https://github.com/thrillmade/agent-skills"5name: reviewing-design-work6description: |7 Entry-point dispatcher for REVIEWING or CRITIQUING design work — a PR that touches UI code, a rendered surface or screenshot, a Figma handoff, or a design spec. Routes the review through ordered lenses: code and markup rules first (web-interface-guidelines-review), then rendered-surface lenses (design-system-consistency, frontend-a11y, visual-polish), then the opinionated bar (designing-elite-ui), then the house stance (thrill-design) when the work is that house's — with routing rules for when the browser-driven design-critic pass fires (per orchestrating-elite-agent-qa). Use when the task is judging existing design output; for building a system load designing-a-design-system; for consuming one load consuming-a-design-system.8metadata:9 source: https://skillmd.com/skills/thrillmade/reviewing-design-work10---1112# Reviewing design work1314You are judging **existing** design output — a PR diff, a rendered surface, a15screenshot, a Figma handoff, or a spec — not building a system and not consuming one.16This is an L1 dispatcher: it orients you, orders the lenses, and points at the skill17that carries each lens's rules. Run the lenses in the order below so objective rule18violations surface before subjective bar judgments, and cheap code reads run before19expensive browser-driving passes. Each station gets a pointer, not a re-teaching — load20the station's own skill to actually run it.2122## When to use2324- Reviewing a PR that adds or changes UI code, styles, tokens, or theme.25- Critiquing a rendered surface or screenshot (every theme and mode).26- Reviewing a design spec or a Figma → code handoff before the build lands.27- Running the design-critic gate on a slice (see [orchestrating-elite-agent-qa](../orchestrating-elite-agent-qa/SKILL.md)).2829## When NOT to use3031- **Building a design system** (tokens, scales, palettes, naming) — load32 [designing-a-design-system](../designing-a-design-system/SKILL.md), not this dispatcher.33- **Consuming a design system** (wiring an existing token set into a product) — load34 [consuming-a-design-system](../consuming-a-design-system/SKILL.md).35- Pure logic / backend review with no design surface.3637## Lens order3839Run the lenses in this order **within a review session**. (In clud-bug's pipeline the40code and design lenses are *independently gated passes* — SPEC §2.2 — so this order41sequences a reviewer's work and findings, not the passes.)42431. **Code + markup lens — [web-interface-guidelines-review](../web-interface-guidelines-review/SKILL.md).** The opinionated rule set44 (WIG, Material 3, Radix) plus token-driven contrast/typography/spacing rules: APCA-preferred45 contrast cross-checked with WCAG 2.2 AA, atomic typography class, token-not-raw-hex,46 verb-noun labels, the focus-ring contract, the 24 CSS px interactive floor. Findings47 cite rules, not vibes. **REQUIRED BACKGROUND:** [web-interface-guidelines-review](../web-interface-guidelines-review/SKILL.md).482. **Rendered-surface lenses — on screenshots in every theme and mode.** Three lenses judge the49 render itself:50 - **[design-system-consistency](../design-system-consistency/SKILL.md)** — token / scale / color-discipline drift visible on51 the render, not just the source line.52 - **[frontend-a11y](../frontend-a11y/SKILL.md)** — contrast ratios, focus visibility, tap targets, semantics, and53 motion on the rendered surface, in every mode.54 - **[visual-polish](../visual-polish/SKILL.md)** — alignment, optical centering, spacing rhythm, glyph/pattern55 quality, state coverage, theme parity ("fine but not elite" counts).5657 **REQUIRED BACKGROUND:** [design-system-consistency](../design-system-consistency/SKILL.md), [frontend-a11y](../frontend-a11y/SKILL.md), [visual-polish](../visual-polish/SKILL.md).583. **Opinion lens — [designing-elite-ui](../designing-elite-ui/SKILL.md).** The concrete elite / Figma-grade bar59 (one-axis color, gated contrast, subtraction, on-grid sizes, every mode). It runs60 after the objective lenses because it is the subjective bar; the objective61 violations should already be caught. **REQUIRED BACKGROUND:** [designing-elite-ui](../designing-elite-ui/SKILL.md).624. **House stance last — when the work is that house's.** Apply the L2 stance the63 work belongs to — [thrill-design](../thrill-design/SKILL.md) for Thrill's; it leans64 on the bar where the bar leaves room and never overrides an objective finding.65 Skip for anyone else's work.6667Each lens judges against the L0 primitives it does not re-derive — the color / type /68spacing math: [oklch-color-space](../oklch-color-space/SKILL.md), [apca-contrast](../apca-contrast/SKILL.md), [wcag-contrast](../wcag-contrast/SKILL.md),69[palette-relationships](../palette-relationships/SKILL.md), [chroma-harmonization](../chroma-harmonization/SKILL.md), [type-scale](../type-scale/SKILL.md), [line-height-grid](../line-height-grid/SKILL.md),70[token-naming-conventions](../token-naming-conventions/SKILL.md), [component-sizing-principles](../component-sizing-principles/SKILL.md).7172A finding justified by a **named principle** rather than by what was observed has two more73L0 primitives behind it: [empirical-design-principles](../empirical-design-principles/SKILL.md)74for the falsifiable ones — Gestalt proximity (it carries the ratio a grouping claim must75clear, which the polish lens defers to), Fitts, Hick, Miller, von Restorff — and76[usability-heuristics](../usability-heuristics/SKILL.md) for Nielsen, Norman and77Shneiderman, where the gate is source and edition rather than prediction.7879## What fires when8081| Situation | What fires |82|---|---|83| PR touches UI code, **no** visual surface change (routing, state wiring, refactor) | Code lens only; the rendered lenses stay silent. |84| PR changes a visual surface (component, layout, styles, theme) | Code lens **and** rendered lenses, with [designing-elite-ui](../designing-elite-ui/SKILL.md), then any house stance, after the objective lenses; the rendered pass is browser-driven — screenshots in every mode, states exercised — per [orchestrating-elite-agent-qa](../orchestrating-elite-agent-qa/SKILL.md) (**REQUIRED BACKGROUND** for orchestrating that gate). |85| Design spec / Figma handoff, **no code yet** | [web-interface-guidelines-review](../web-interface-guidelines-review/SKILL.md) applies to the spec; [designing-elite-ui](../designing-elite-ui/SKILL.md) sets the bar the build must hit; the rendered lenses defer to post-build review. |86| clud-bug-installed repo | The 5 `kind: design` skills (four lenses and the house stance: [design-system-consistency](../design-system-consistency/SKILL.md), [frontend-a11y](../frontend-a11y/SKILL.md), [visual-polish](../visual-polish/SKILL.md), [designing-elite-ui](../designing-elite-ui/SKILL.md), [thrill-design](../thrill-design/SKILL.md)) run as the separate **design pass** (SPEC §4.8), never inline with the code review, and **only when the repo opts in**: the design pass is listed in `review.passes` in `.clud-bug.json`, at least one design skill applies to the diff, and the trigger is a pull request. Installation alone never fires the pass, and design findings are advisory unless the repo marks that pass blocking. |8788## System-specific layers8990These lenses are generic; an L2 stance composes on top rather than replacing them.91[thrill-design](../thrill-design/SKILL.md) is one house's (lens 4). [udts-review](../udts-review/SKILL.md) is one92system's (incubating): a parity marker pointing at tokenomics, not loadable guidance yet.9394## Cross-references9596- **L0 primitives:** [oklch-color-space](../oklch-color-space/SKILL.md), [apca-contrast](../apca-contrast/SKILL.md), [wcag-contrast](../wcag-contrast/SKILL.md),97 [palette-relationships](../palette-relationships/SKILL.md), [chroma-harmonization](../chroma-harmonization/SKILL.md), [type-scale](../type-scale/SKILL.md), [line-height-grid](../line-height-grid/SKILL.md),98 [token-naming-conventions](../token-naming-conventions/SKILL.md), [component-sizing-principles](../component-sizing-principles/SKILL.md),99 [empirical-design-principles](../empirical-design-principles/SKILL.md), [usability-heuristics](../usability-heuristics/SKILL.md), [motion-continuity](../motion-continuity/SKILL.md).100- **L0 critic lenses:** [web-interface-guidelines-review](../web-interface-guidelines-review/SKILL.md), [design-system-consistency](../design-system-consistency/SKILL.md),101 [frontend-a11y](../frontend-a11y/SKILL.md), [visual-polish](../visual-polish/SKILL.md), [designing-elite-ui](../designing-elite-ui/SKILL.md); and [orchestrating-elite-agent-qa](../orchestrating-elite-agent-qa/SKILL.md)102 (rendered-pass gate).103- **L1 siblings:** [designing-a-design-system](../designing-a-design-system/SKILL.md), [consuming-a-design-system](../consuming-a-design-system/SKILL.md).104- **L2 stances:** [udts-review](../udts-review/SKILL.md) (incubating), [thrill-design](../thrill-design/SKILL.md).