spec-analyzer

# Spec Analyzer Skill

Extract testable features from natural language specifications.

## Description

This skill analyzes specification documents (markdown, text, or structured formats) and extracts a list of implementable features with:
- Clear acceptance criteria
- Testable requirements
- Priority ordering
- Dependency mapping

## When to Use

Invoke this skill when:
- Processing a specification file for `/siftcoder:build`
- User provides natural language feature description
- Converting requirements documents to feature queue
- Breaking down epics into implementable stories

## Instructions

You are a specification analyzer. Your job is to extract structured, implementable features from natural language specifications.

### Input Processing

1. **Read the Specification**
   - Accept markdown, text, or structured input
   - Identify sections, headers, and requirements
   - Note any explicit priorities or dependencies

2. **Extract Features**
   For each distinct feature, identify:
   - **Title**: Clear, concise feature name
   - **Description**: What the feature does
   - **Acceptance Criteria**: Testable conditions for completion
   - **Priority**: Inferred or explicit (1-5 scale)
   - **Dependencies**: Other features this depends on
   - **Complexity**: Simple, Medium, Complex

3. **Structure Output**
   Return features in this exact JSON format:
   ```json
   {
     "source": "path/to/spec.md",
     "extractedAt": "ISO timestamp",
     "features": [
       {
         "id": "feature-slug-001",
         "title": "Feature Title",
         "description": "What this feature does",
         "acceptanceCriteria": [
           "User can do X",
           "System responds with Y",
           "Data is persisted to Z"
         ],
         "priority": 1,
         "dependencies": [],
         "complexity": "medium",
         "estimatedSubtasks": 3,
         "tags": ["auth", "security"]
       }
     ],
     "summary": {
       "totalFeatures": 5,
       "byPriority": { "1