本文へ移動
cccskills
無料GitHub で公開

ui-designer

Extract design systems from reference UI images and generate implementation-ready UI design prompts. Use when users provide UI screenshots/mockups and want to create consistent designs, generate design systems, or build MVP UIs matching reference aesthetics.

インストール方法を見る

含まれるファイル(4)

  • SKILL.md6.8 KB
  • assets/app-overview-generator.md2.1 KB
  • assets/design-system.md4.0 KB
  • assets/vibe-design-template.md2.1 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

UI Designer

Overview

This skill enables systematic extraction of design systems from reference UI images through a multi-step workflow: analyze visual patterns → generate design system documentation → create PRD → produce implementation-ready UI prompts.

When to Use

  • User provides UI screenshots, mockups, or design references
  • Need to extract color palettes, typography, spacing from existing designs
  • Want to generate design system documentation from visual examples
  • Building MVP UI that should match reference aesthetics
  • Creating multiple UI variations following consistent design principles

Workflow

Step 1: Gather Inputs

Request from user:

  • Reference images directory: Path to folder containing UI screenshots/mockups
  • Project idea file: Document describing the product concept and goals
  • Existing PRD (optional): If PRD already exists, skip Step 3

Step 2: Extract Design System from Images

Use Task tool with general-purpose subagent, providing:

Prompt template from assets/design-system.md:

  • Analyze color palettes (primary, secondary, accent, functional colors)
  • Extract typography (font families, sizes, weights, line heights)
  • Identify component styles (buttons, cards, inputs, icons)
  • Document spacing system
  • Note animations/transitions patterns
  • Include dark mode variants if present

Attach reference images to the subagent context.

Output: Complete design system markdown following the template format

Save to: documents/designs/{image_dir_name}_design_system.md

Step 3: Generate MVP PRD (if not provided)

Use Task tool with general-purpose subagent, providing:

Prompt template from assets/app-overview-generator.md:

  • Replace {项目背景} with content from project idea file
  • The template guides through: elevator pitch, problem statement, target audience, USP, features list, UX/UI considerations

Interact with user to refine and clarify product requirements

Output: Structured PRD markdown

Save as variable for Step 4 (optionally save to documents/prd/)

Step 4: Compose Final UI Implementation Prompt

Combine design system and PRD using assets/vibe-design-template.md:

Substitutions:

  • {项目设计指南} → Design system from Step 2
  • {项目MVP PRD} → PRD from Step 3 or provided PRD file

Result: Complete, implementation-ready prompt containing:

  • Design aesthetics principles
  • Project-specific color/typography guidelines
  • App overview and feature requirements
  • Implementation tasks (multiple UI variations, component structure)

Save to: documents/ux-design/{idea_file_name}_design_prompt_{timestamp}.md

Step 5: Verify React Environment

Check for existing React project:

find . -name "package.json" -exec grep -l "react" {} \;

If none found, inform user:

npx create-react-app my-app
cd my-app
npm install -D tailwindcss postcss autoprefixer
npx tailwindcss init -p
npm install lucide-react

Step 6: Implement UI

Use the final composed prompt from Step 4 to implement UI in React project.

The prompt instructs to:

  • Create multiple design variations (3 for mobile, 2 for web)
  • Organize as separate components: [solution-name]/pages/[page-name].jsx
  • Aggregate all variations in showcase page

Template Assets

assets/design-system.md

Template for extracting visual design patterns. Includes sections for:

  • Color palette (primary, secondary, accent, functional, backgrounds)
  • Typography (font families, weights, text styles)
  • Component styles (buttons, cards, inputs, icons)
  • Spacing system (4dp-48dp scale)
  • Animations (durations, easing curves)
  • Dark mode variants

Use this template when analyzing reference images to ensure comprehensive design system coverage.

assets/app-overview-generator.md

Template for collaborative PRD generation. Guides through:

  • Elevator pitch
  • Problem statement and target audience
  • Unique selling proposition
  • Platform targets
  • Feature list with user stories
  • UX/UI considerations per screen

Designed for interactive refinement with user to clarify requirements.

assets/vibe-design-template.md

Final implementation prompt template combining design system and PRD. Includes:

  • Aesthetic principles (minimalism, whitespace, color theory, typography hierarchy)
  • Practical requirements (Tailwind CSS, Lucide icons, responsive design)
  • Task specifications (multiple variations, component organization)

This template produces prompts ready for UI implementation without further modification.

Best Practices

Image Analysis

  • Read all images before starting analysis
  • Look for patterns across multiple screens
  • Note both explicit styles (colors, fonts) and implicit principles (spacing, hierarchy)
  • Capture dark mode if present in references

Design System Extraction

  • Be systematic: cover all template sections
  • Use specific values (hex codes, px sizes) not generic descriptions
  • Document the "why" for design choices when inferable
  • Include variants (hover states, disabled states)

PRD Generation

  • Engage user interactively to clarify ambiguities
  • Suggest features based on problem understanding
  • Ensure MVP scope is realistic
  • Document UX considerations per screen/interaction

Output Organization

  • Save design system with descriptive filename (based on image dir name)
  • Save final prompt with timestamp for version tracking
  • Keep all outputs in documents/ directory for easy reference
  • Preserve intermediate outputs for iteration

Example Usage

User provides:

  • reference-images/saas-dashboard/ (5 screenshots)
  • ideas/project-management-app.md (project concept)

Execute workflow:

  1. Read 5 images from reference-images/saas-dashboard/
  2. Use Task tool → design-system.md template → analyze images
  3. Save to documents/designs/saas-dashboard_design_system.md
  4. Use Task tool → app-overview-generator.md with project concept
  5. Refine PRD through user interaction
  6. Combine design system + PRD using vibe-design-template.md
  7. Save to documents/ux-design/project-management-app_design_prompt_20251025_153000.md
  8. Check React environment, inform user if setup needed
  9. Implement UI using final prompt

Notes

  • This is a high freedom workflow—adapt steps based on context
  • Templates provide structure but encourage thoughtful analysis over rote filling
  • User interaction during PRD generation is critical for quality
  • Final prompt quality directly impacts UI implementation success
  • Preserve all intermediate outputs for iteration and refinement

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Fixes web search on an agent whose model backend can't run it: a relay/reseller proxying Claude or Codex returns empty instead of failing. Use when web search returns nothing, a model insists a shipped product doesn't exist, someone wants to give an agent internet access, or the user is on a third-party base URL, relay, or 中转站. Diagnoses which built-in tools are dead, removes them, and installs a working replacement.

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

抓取 A 股消息面情报:从财联社、华尔街见闻、金十、新浪 7x24、东财快讯、 证监会/央行/上交所/财政部政策公告、东方财富股吧等公开来源抓取与股票相关的 新闻、政策、情绪,输出结构化 JSON 或 Markdown。 当用户提到“A 股消息面”、“抓新闻”、“个股消息”、“政策监管”、“股吧情绪”、 “财联社”、“东财快讯”、“市场情绪”或需要把某只股票相关的公开情报聚合出来时 触发。也适用于“帮我看看 000001 最近有什么消息”这类口语化请求。

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

Transcribes audio or video to speaker-labeled, timestamped text, locally with MLX on Apple Silicon or remotely. Use for 转录 / 录音转文字 / 说话人分离 / 字幕, and also for preparing audio for ASR without transcribing: 转格式, 降采样到 16kHz, merging recorder segments, or compressing and speeding up audio before 飞书妙记 — even when it looks like a one-line ffmpeg job.

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

Routes audio: StepFun ASR/语音识别, StepFun TTS/配音, transcript/妙记→会议纪要, merge/review minutes. Reads one bundled specialist; generic ASR and correction stay direct.

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

Diagnoses and repairs repository setup and guarded Git workflows for Claude Code or Codex — environment repair, startup sync, hook auditing, collaborator handoff. Use when a repo won't run, a teammate onboards, hook output duplicates, or commit/push/conflict needs guarding. Not for lost-commit recovery (use git-safety-net), GitHub ops (use github-ops), or history scrubbing (use github-sensitive-data-cleanup).

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

Runs adversarial due-diligence on a benchmark the user envies — a founder, KOL, company, or product whose success looks inflated — splitting marketing bubble from real signal, then mapping the validated playbook onto the user's own resources. Use for 尽调/对标/拆解 a competitor, 抄/偷师 their playbook, or suspecting 水分/泡沫 in claims. Prefer over deep-research when debunking inflated claims, not a neutral briefing.

日本語の概要は準備中です。原文の説明を表示しています。

daymade/claude-code-skills1,4512026年10月11日 更新

daymade のスキルをすべて見る

このスキルの問題を報告する