Concept Extractor Skill

Concept Extractor Skill

코드를 분석하여 학습할 만한 프로그래밍 개념을 추출합니다.

When to Use

• PostToolUse 훅에서 자동 호출 (백그라운드)
• /learn-capture 커맨드에서 현재 컨텍스트 분석
• 사용자가 "이 코드에서 무엇을 배울 수 있나요?" 질문 시

Extraction Strategy

1. Language Features (언어 기능)

Python:

• Decorators (@property, @staticmethod)
• Context managers (with statement)
• List/dict comprehensions
• Generators and yield
• Type hints (typing module)
• Async/await patterns

JavaScript/TypeScript:

• Arrow functions and closures
• Promises and async/await
• Destructuring
• Spread/rest operators
• Generics (TypeScript)
• Type guards (TypeScript)

Other languages: Similar language-specific features

2. Framework/Library APIs

Web Frameworks:

• React: Hooks (useState, useEffect, custom hooks)
• Vue: Composition API, reactivity
• Flask/Django: Routing, ORM, middleware
• Express: Middleware, routing

UI Frameworks:

• PyQt: Signals/slots, layouts, widgets
• Tkinter: Event handling, geometry managers

Data/ML:

• Pandas: DataFrame operations
• NumPy: Array broadcasting
• TensorFlow/PyQt: Model definition patterns

3. Design Patterns

• Creational: Factory, Singleton, Builder
• Structural: Adapter, Decorator, Facade
• Behavioral: Observer, Strategy, Command

4. Algorithms & Data Structures

• Sorting algorithms
• Search algorithms (binary search, DFS, BFS)
• Dynamic programming patterns
• Tree/graph traversal
• Hash tables, heaps, tries

5. Best Practices

• Error handling patterns
• Testing patterns (mocking, fixtures)
• Code organization (modules, packages)
• API design principles

Extraction Rules

Include:

✅ Specific and learnable concepts:

• "React useEffect cleanup functions"
• "Python context managers"
• "Factory pattern with dependency injection"

✅ Non-trivial patterns:

• "Async error handling with try/catch"
• "Memoization for performance optimization"

Exclude:

❌ Too basic concepts:

• "Variables"
• "Functions"
• "If statements"

❌ Too broad concepts:

• "Python"
• "Web development"
• "Programming"

❌ Project-specific details:

• "UserController" (unless demonstrating a pattern)
• "calculateTotal function" (unless algorithmic)

Output Format

Return a structured list:

{
  "concepts": [
    {
      "name": "React useEffect Cleanup",
      "category": "framework-api",
      "language": "javascript",
      "framework": "react",
      "difficulty": "intermediate",
      "description": "Cleanup functions in useEffect to prevent memory leaks",
      "code_snippet": "[relevant code]",
      "why_important": "Prevents memory leaks and stale closures"
    }
  ],
  "context": {
    "file": "[file path]",
    "lines": "[line range]",
    "project_context": "[what the code does]"
  }
}

Example

Input Code

from functools import lru_cache

@lru_cache(maxsize=128)
def fibonacci(n):
    if n < 2:
        return n
    return fibonacci(n-1) + fibonacci(n-2)

Output

{
  "concepts": [
    {
      "name": "Python lru_cache Decorator",
      "category": "language-feature",
      "language": "python",
      "difficulty": "intermediate",
      "description": "Memoization decorator from functools for caching function results",
      "why_important": "Dramatically improves performance for recursive functions"
    },
    {
      "name": "Memoization Pattern",
      "category": "algorithm-technique",
      "difficulty": "intermediate",
      "description": "Caching expensive function calls to avoid recomputation",
      "why_important": "Common optimization technique in dynamic programming"
    }
  ]
}

Integration with Learning System

1. Deduplication: Check if concept already tracked today
2. Enrichment: Add file context and timestamp
3. Storage: Append to .claude/learning/concepts.json
4. Notification: (Only if user explicitly requests analysis)