Claude Code SubAgents

"Pokemon for Developers" - Your AI agents level up, form teams, and tackle any coding challenge

Quick Start: cd ~/.claude && git clone https://github.com/mylee04/claude-code-subagents.git agents (10 seconds)

What Is This?

A gamified collection of 31 specialized AI agents that work together like a development team. Each agent levels up as you use them, unlocks achievements, and forms powerful combinations for complex tasks.

Core Value: Instead of one generic AI assistant, you get a specialized team where each agent is an expert in their domain (Python, DevOps, Security, etc.).

Key Features

• 31 Specialist Agents - From Python experts to security auditors
• Gamification System - Agents gain XP, level up (Novice → Legend), earn achievements
• Smart Agent Registry - Automatically discovers and coordinates all available agents
• Team Formations - Agents work together with synergy bonuses
• Real Progress Tracking - Watch your AI team collaborate in real-time

Interactive Demo

Try it now - see the XP system, achievements, and agent collaboration in action:

cd ~/.claude/agents && python3 gamification/scripts/agents-demo.py

Watch agents level up, earn achievements like "Speed Demon" and "Bug Hunter", and see real-time mission progress.

Basic Usage

# Use specific agents
/python-pro optimize this function
/devops-troubleshooter analyze these logs
/security-auditor review authentication flow

# Complex multi-agent tasks
# Use Claude Code's built-in orchestration for complex workflows

🔍 Agent Discovery & Coordination System

The SubAgents system now features an advanced AgentRegistry that automatically discovers and coordinates agents from multiple locations:

Live Demo - See It In Action!

🔍 AGENT DISCOVERY
─────────────────
✓ Successfully discovered 34 agents

  ▸ Discovery Statistics
  Total Agents Found....... 34
  Categories Discovered.... 8
  Tech Stacks Available.... 11

📊 CATEGORY BREAKDOWN
───────────────────────

  Development         │████████████████████│ 10 agents
  Infrastructure      │████████████        │  6 agents
  Data & AI           │████████            │  4 agents
  Quality Assurance   │████████            │  4 agents
  Product             │████████            │  4 agents
  Coordination        │██████              │  3 agents
  Business            │████                │  2 agents
  Security            │██                  │  1 agents

🎯 FEATURE PLANNING & TEAM RECOMMENDATION
────────────────────────────────────────

  ▸ Building E-Commerce Web Application
  📝 Tech Stack: ["react", "typescript", "python", "postgresql"]
  
  ✓ Recommended Team (6 agents):
  • full-stack-architect - Complete system architecture design
  • python-elite - Python/Django/FastAPI development
  • frontend-developer - Modern UI components and frameworks
  • database-optimizer - Database performance specialist
  • devops-engineer - Full-stack DevOps specialist
  • security-auditor - Security vulnerability scanner

Run the Demo Yourself

# Run the colorful interactive demo
python3 demo_agent_registry.py

# Or try the simple test
python3 test_agent_registry.py

Key Features

• Unified Discovery: Finds agents from /agents/, ~/.claude/agents/, and .claude/agents/
• Built-in Orchestration: Claude Code handles agent coordination automatically
• Team Formation: Intelligently recommends 3-6 agents based on your project needs
• No External Dependencies: Works with built-in Python libraries only

📋 All 31 Agents (Click to expand)

Command & Planning

• Use Claude Code's built-in orchestration capabilities for multi-agent workflows

Development

• python-pro - Python/Django/FastAPI expert
• javascript-pro - JavaScript/TypeScript/React specialist
• golang-pro - Go language expert
• rust-pro - Rust programming specialist
• sql-pro - Database query optimization
• backend-architect - API design and microservices
• frontend-developer - Modern UI components and frameworks
• full-stack-architect - Complete system design

Infrastructure & DevOps

• devops-engineer - CI/CD and containerization
• devops-troubleshooter - Production incident response
• cloud-architect - AWS/GCP/Azure infrastructure
• deployment-engineer - Release automation
• database-optimizer - Performance tuning
• incident-commander - Emergency response coordination

Quality & Security

• code-reviewer - Code quality and best practices
• test-engineer - Testing strategy and implementation
• performance-engineer - Optimization and profiling
• quality-engineer - QA processes and standards
• security-auditor - Vulnerability scanning and compliance

Data & AI

• data-engineer - ETL pipelines and data architecture
• ai-engineer - LLM integration and RAG systems
• ml-engineer - Model training and deployment
• data-ai-ml-engineer - Full-stack data science

Product & Business

• api-documenter - Documentation and guides
• dx-optimizer - Developer experience improvement
• user-feedback-analyst - User research and insights
• business-analyst - Metrics and KPI analysis
• content-marketer - Content strategy and SEO
• tech-portfolio-reviewer - Resume and portfolio optimization

Tools

• git-specialist - Version control operations

🎮 How the Gamification Works

Agent Levels & Tiers

• 🟢 Novice (Lv.1-10) - Learning the basics, understanding your patterns
• 🔵 Adept (Lv.11-30) - Developing expertise, understanding project conventions
• 🟡 Expert (Lv.31-70) - Mastering techniques, anticipating your needs
• 🟠 Master (Lv.71-120) - Elite performance, proactive recommendations
• 🔴 Grandmaster (Lv.121-200) - Legendary status, deep project insight
• 💎 Legend (Lv.201+) - Ultimate mastery, transcenden