meta-architect-claudecode

# Meta-Architect

Choose the right customization layer and build effective skills for Claude Code using official documentation.

## 🚨 MANDATORY: Read BEFORE Proceeding

**CRITICAL**: You MUST read and understand these URLs:

### Primary Documentation (MUST READ)
- **[MUST READ] Official Skills Guide**: https://code.claude.com/docs/en/skills
  - **Tool**: `mcp__simplewebfetch__simpleWebFetch`
  - **When to Read**: Before any customization layer decisions
  - **Content**: Skills architecture, progressive disclosure, best practices
  - **Cache**: 15 minutes minimum

- **[MUST READ] Subagents Guide**: https://code.claude.com/docs/en/sub-agents
  - **Tool**: `mcp__simplewebfetch__simpleWebFetch`
  - **When to Read**: When considering subagent usage
  - **Content**: Agent coordination patterns, state management, context: fork
  - **Cache**: 15 minutes minimum

- **[MUST READ] Plugin Architecture**: https://code.claude.com/docs/en/plugins
  - **Tool**: `mcp__simplewebfetch__simpleWebFetch`
  - **When to Read**: Before plugin structure decisions
  - **Content**: Plugin organization, component architecture
  - **Cache**: 15 minutes minimum

### ⚠️ BLOCKING RULES
- **DO NOT proceed** until you've fetched and reviewed Primary Documentation
- **REQUIRED** to understand skills-first architecture before layer selection
- **REQUIRED** to understand progressive disclosure before implementation

---

## Quick Decision Tree

### Layer Selection (START HERE)

```
START: What do you need?
│
├─ "Persistent project norms"
│  └─→ CLAUDE.md rules
│
├─ "Domain expertise to discover"
│  └─→ Skill (Minimal Pack: CLAUDE.md + Skill)
│     ├─ "Simple, focused task" → Skill (regular)
│     └─ "Complex workflow needing isolation" → Skill (context: fork)
│
├─ "Event automation"
│  └─→ Hook
│
├─ "Service integration"
│  └─→ MCP
│
└─ "Isolation/parallelism"
   └─→ Subagent (RARE/ADVANCED)
```

**Context: Fork Triggers:**
- High-volume output (extensive grep, repo traversal)
- Noisy exploration that