screenshot-feature-extractor
Extract features from UI screenshots
Analyzing screenshots manually to extract features and create task lists is time-consuming and error-prone. This multi-agent skill analyzes UI screenshots in parallel from three perspectives to identify features, interactions, and business logic, then generates a ready-to-use development task checklist.
์คํฌ ZIP ๋ค์ด๋ก๋
Claude์์ ์ ๋ก๋
์ค์ โ ๊ธฐ๋ฅ โ ์คํฌ โ ์คํฌ ์ ๋ก๋๋ก ์ด๋
ํ ๊ธ์ ์ผ๊ณ ์ฌ์ฉ ์์
ํ ์คํธํด ๋ณด๊ธฐ
"screenshot-feature-extractor" ์ฌ์ฉ ์ค์ ๋๋ค. Analyze this dashboard screenshot and create a feature task list
์์ ๊ฒฐ๊ณผ:
- User Authentication module: Login screen, Remember me, Password reset
- Dashboard Overview module: Stats cards, Real-time data, Date range picker
- Navigation module: Sidebar menu, Breadcrumbs, User menu dropdown
- Reporting module: Export PDF, Date filter, Graph visualization
"screenshot-feature-extractor" ์ฌ์ฉ ์ค์ ๋๋ค. Analyze these 5 screenshots from an e-commerce app
์์ ๊ฒฐ๊ณผ:
- Product Listing: Search bar, Category filters, Sort options, Product cards with images and prices
- Product Detail: Image gallery, Add to cart, Size selector, Reviews section
- Cart: Quantity controls, Total calculation, Promo code input, Checkout button
- Checkout: Shipping form, Payment options, Order summary, Confirm purchase
"screenshot-feature-extractor" ์ฌ์ฉ ์ค์ ๋๋ค. Compare this competitor screenshot against our similar feature
์์ ๊ฒฐ๊ณผ:
- Competitor unique features: Dark mode, Voice search, Social sharing integration
- Feature gaps: Missing in our product - Collaborative playlists, Offline mode
- Priority recommendations: Implement voice search first, then dark mode, then social features
๋ณด์ ๊ฐ์ฌ
์์ This skill is pure documentation with no executable code. It only contains markdown prompts that orchestrate AI agents. No Ruby, shell, Python, JavaScript, or TypeScript files exist. All static findings are false positives from the scanner misinterpreting markdown backticks and JSON structures. The skill reads user-provided screenshots and writes output to predictable local paths.
์ํ ์์ธ
โ๏ธ ์ธ๋ถ ๋ช ๋ น์ด (14)
๐ ๋คํธ์ํฌ ์ ๊ทผ (1)
ํ์ง ์ ์
๋ง๋ค ์ ์๋ ๊ฒ
Create requirements from mockups
Turn UI mockups or screenshots into feature specs and task lists for engineering teams
Understand feature scope quickly
Analyze competitor apps or reference designs to understand feature scope before implementation
Document existing products
Reverse-engineer existing products by analyzing their screens to document capabilities
์ด ํ๋กฌํํธ๋ฅผ ์ฌ์ฉํด ๋ณด์ธ์
Analyze this screenshot and extract all visible features, UI components, and interactions. Output file: docs/plans/YYYY-MM-DD-app-features.md
Analyze these screenshots from the same application. Identify all features across screens and create a consolidated task list. Deduplicate similar features.
Analyze these competitor screenshots. Identify unique features and gaps compared to our product. Create a feature comparison matrix with development task priorities.
Perform deep analysis on this screenshot. Break down each feature into 3-5 granular subtasks focusing on user interactions and business logic. Output detailed development checklist.
๋ชจ๋ฒ ์ฌ๋ก
- Provide context about the product purpose and target users for better analysis
- Use high-resolution screenshots with clear text for accurate feature detection
- Include multiple related screens to capture navigation and state transitions
ํผํ๊ธฐ
- Using low-resolution or blurry screenshots that obscure UI details
- Analyzing screenshots without providing product context or user goals
- Expecting implementation-level details like tech stack or database schema
์์ฃผ ๋ฌป๋ ์ง๋ฌธ
What image formats are supported?
How many screenshots can I analyze at once?
Can I integrate this with my design tools?
Is my data sent to external servers?
Why are my screenshots not detecting features accurately?
How does this compare to manual feature mapping?
๊ฐ๋ฐ์ ์ธ๋ถ ์ ๋ณด
์์ฑ์
notedit๋ผ์ด์ ์ค
MIT
๋ฆฌํฌ์งํ ๋ฆฌ
https://github.com/notedit/happy-coding-agent/tree/main/skills/screenshot-feature-extractor์ฐธ์กฐ
main
ํ์ผ ๊ตฌ์กฐ