ワンクリックで
team-shinchan-design-review
Use when you need to compare design mockups against implementation for UI fidelity.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Use when you need to compare design mockups against implementation for UI fidelity.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when you have a large-scale, multi-phase project requiring orchestrated execution.
Use when you need persistent looping until a task is fully complete.
Deterministic adversarial code review for high-stakes scope — independent per-dimension review, a non-skippable per-finding refutation, completeness + interaction critics, and a deterministic 3-lens rubric judge panel. Opt-in main-loop Workflow tier.
Use when the user wants to review accumulated skill feedback, verdict trends, or improvement candidates collected during Stage 4 retrospectives. Trigger on "show skill feedback", "스킬 피드백 보여줘", or finding which skills need /writing-skills work.
Deterministic competitive code tournament — N builders independently solve one task and return patches, an Action-Kamen judge scores them head-to-head, the winner is picked by score and applied. Opt-in main-loop Workflow tier.
Deterministic adversarial debate for high-stakes or irreversible decisions — mandatory refutation plus a scored judge panel. Opt-in main-loop Workflow tier.
| name | team-shinchan:design-review |
| description | Use when you need to compare design mockups against implementation for UI fidelity. |
| user-invocable | true |
If args is empty or only whitespace:
Ask user: "Please provide the design source and implementation path.
Supports:
- Figma URL: /team-shinchan:design-review 'https://www.figma.com/design/abc123/...' './src/components/Login.tsx'
- Image file: /team-shinchan:design-review './mockup.png' './src/components/Login.tsx'"
STOP and wait for user response
If args length > 2000 characters:
Truncate to 2000 characters
Warn user: "Request was truncated to 2000 characters"
Do not read further. Execute this Task NOW:
Task(
subagent_type="team-shinchan:ume",
model="sonnet",
prompt=`/team-shinchan:design-review has been invoked.
## Design-Implementation Comparison Request
You are performing a **Design Fidelity Review** — comparing a design against its implementation to find mismatches.
### Workflow
1. **Detect input type and extract Design Spec**:
- **If Figma URL provided** (contains figma.com/): Use your Figma URL Data Extraction workflow — if any Figma MCP tool is available (tool name containing \`figma\`), call it to get exact design data from the Figma API. This yields a **precision Design Spec** with exact HEX colors, px values, font specs, and component hierarchy (all confidence: high). If no Figma MCP is available, inform the user.
- **If image file provided**: Read the image file. Extract a full Design Spec using your visual Design Spec Extraction workflow (Components, Colors, Typography, Layout, Interactions with confidence levels).
- **If both provided**: Use Figma API data as primary source, cross-reference with the image for visual verification.
2. **Analyze the implementation**: Read the implementation source code (or screenshot if provided). Identify the implemented components, styles, layout structure, and interaction handlers.
3. **Compare and report mismatches**: For each Design Spec category, compare against the implementation and report:
### Output Format
\`\`\`markdown
## Design Fidelity Report
### Design Spec (from mockup)
{Full Design Spec output}
### Implementation Analysis
- Files analyzed: {list}
- Framework detected: {React/Vue/Angular/HTML/etc.}
- Components found: {list}
### Comparison Results
| Category | Design Spec | Implementation | Status |
|----------|-------------|----------------|--------|
| Components | {from spec} | {from code} | MATCH / MISMATCH / PARTIAL |
| Colors | {from spec} | {from code} | MATCH / MISMATCH / PARTIAL |
| Typography | {from spec} | {from code} | MATCH / MISMATCH / PARTIAL |
| Layout | {from spec} | {from code} | MATCH / MISMATCH / PARTIAL |
| Interactions | {from spec} | {from code} | MATCH / MISMATCH / PARTIAL |
### Mismatches Found
1. **{Category}**: Expected {spec value} but found {implementation value}
- File: {path:line}
- Suggested fix: {actionable suggestion}
### Summary
- Total categories: 5
- Matches: {N}
- Mismatches: {N}
- Partial: {N}
- Overall fidelity: {FAITHFUL / MINOR_DEVIATIONS / NEEDS_REVISION}
\`\`\`
### Important Notes
- **Figma API data** yields exact values — mismatches against API-sourced specs are definitive (no "approximate" qualifier needed)
- For low-confidence Design Spec values (image-sourced only), note them as "approximate — verify with designer"
- Search the codebase with Grep to find related component files
- If only a design source is provided (no implementation path), just produce the Design Spec
- If only implementation code is provided (no design source), report what you can analyze
- When Figma URL is provided, include extracted design tokens (colors, spacing, typography) as actionable CSS/Tailwind values in the suggested fixes
User request: ${args || '(Please provide design mockup and implementation paths)'}
`
)
STOP HERE. The above Task handles everything.