A comprehensive collection of production-grade Claude Code configurations, specialized agents, and automation workflows for optimizing AI-assisted development

This repository serves as an educational reference for developers looking to maximize their productivity with Claude Code. It demonstrates advanced configuration patterns, custom agents, and automation workflows that have been refined through extensive real-world usage in building production AI systems, MCP servers, and vector databases.

⚠️ Note: This is a reference repository intended for learning. All credentials shown are placeholders. Adapt these patterns to your specific needs and environment.

• 🤖 Create specialized AI agents for domain-specific expertise
• ⚙️ Configure advanced permissions and security controls
• 🔧 Build custom commands for complex workflows
• 🚀 Automate repetitive tasks with intelligent hooks
• 💾 Implement memory systems with vector search
• 🏗️ Deploy production MCP servers with monitoring
• 🔍 Debug and optimize AI-assisted development

.claude/
├── agents/          # 15 specialized AI agents (~15,000 lines)
├── commands/        # 7 custom workflow commands
├── hooks/          # 2 automation scripts
└── settings.json   # Central configuration hub

This configuration includes 15 expert agents, each with deep domain knowledge and specific capabilities:

• code-reviewer - Comprehensive code review with security focus
• debugger - Systematic debugging and root cause analysis
• test-runner - Automated testing with MCP protocol validation

• mcp-protocol-expert - Protocol debugging and compliance validation
• mcp-sdk-builder - SDK implementation patterns
• mcp-transport-expert - Transport layers (stdio, HTTP, SSE, WebSocket)
• mcp-types-expert - TypeScript type system and Zod schemas

• neon-drizzle-expert - Neon PostgreSQL with Drizzle ORM
• pgvector-advanced - Advanced pgvector v0.8.0 features
• vector-search-expert - Semantic search and embeddings

• memory-architecture - Database design and indexing strategies
• memory-lifecycle - Consolidation and expiration management
• memory-validator - Data integrity and validation
• companion-architecture - Multi-tenant AI companion systems
• production-deployment - HTTPS deployment with Kubernetes

Streamline complex workflows with these custom commands:

/setup quick       # Quick project setup
/setup full        # Complete environment with all dependencies
/setup database    # Database-focused initialization

/test             # Generate comprehensive test suites
/review           # Security-focused code review
/explain          # Context-aware code explanation

/mcp-debug        # Debug MCP protocol issues
/memory-ops       # Test memory CRUD operations
/perf-monitor     # Performance profiling

Automatically triggered on file modifications:

• ✅ Type checking with tsc --noEmit
• ✨ Prettier formatting
• 🔧 ESLint fixing
• 🧪 Test execution for test files
• 📁 Smart filtering (skips node_modules, build dirs)

• 📝 Logs all executed commands for audit trails
• ⏱️ Timestamps for debugging

{
  "permissions": {
    "allow": [
      "Read", "Grep", "Glob", "LS",
      "Bash(npm test:*)",
      "Write(**/*.ts)"
    ],
    "deny": [
      "Read(./.env)",
      "Bash(rm -rf:*)",
      "Bash(git push:*)"
    ]
  }
}

• 🔒 Whitelist approach for commands
• 🚫 Prevents access to sensitive files
• ✅ Scoped write permissions
• 🛡️ Git operation safety

Use the memory-focused agents to:

• Design vector-indexed database schemas
• Implement semantic search with pgvector
• Handle memory lifecycle and consolidation
• Deploy with production monitoring

Leverage deployment agents for:

• Docker containerization with security hardening
• Kubernetes orchestration with auto-scaling
• Prometheus/Grafana monitoring setup
• Structured logging and distributed tracing

Build multi-tenant companion systems with:

• Isolated memory spaces per user
• Token-based authentication
• Rate limiting and quota management
• Conversation history with vector search

# Explore the agents
ls -la .claude/agents/

# Review custom commands
cat .claude/commands/setup.md

# Examine hooks
cat .claude/hooks/typescript-dev.sh

1. Choose relevant agents for your domain
2. Customize commands for your tech stack
3. Configure permissions based on your security needs
4. Set up hooks for your development workflow

"env": {
  "DATABASE_URL": "postgresql://user:pass@host/dbname?sslmode=require",
  "OPENAI_API_KEY": "sk-your-openai-api-key-here",
  "MCP_SERVER_PORT": "3000"
}

"hooks": {
  "PostToolUse": [{
    "matcher": "Edit|MultiEdit|Write",
    "hooks": [{
      "type": "command",
      "command": "$CLAUDE_PROJECT_DIR/.claude/hooks/typescript-dev.sh"
    }]
  }]
}

• 15 Specialized Agents - ~15,000 lines of domain expertise
• 7 Custom Commands - Covering full development lifecycle
• 2 Automation Hooks - TypeScript and logging automation
• 109 Lines of Settings - Fine-tuned configuration
• Version-Locked Stack - Production-tested dependencies

This configuration is optimized for:

• TypeScript & Node.js development
• PostgreSQL 17 with Neon serverless
• Drizzle ORM v0.44.4 for type-safe database access
• pgvector v0.8.0 for vector similarity search
• MCP SDK for protocol implementation
• Docker & Kubernetes for deployment

1. 🚀 10x Faster Development - Specialized agents handle complex tasks
2. 🛡️ Enhanced Security - Multi-layered permission controls
3. 🤖 Automated Workflows - Hooks handle routine tasks
4. 📚 Built-in Expertise - Deep domain knowledge in agents
5. 🔧 Production-Ready - Deployment and monitoring included
6. 🧪 Quality Assurance - Automated testing and review

While this is primarily a reference repository, suggestions for improvements are welcome. Consider:

• Additional agent specializations
• New workflow commands
• Enhanced automation hooks
• Security improvements

This configuration reference is provided as-is for educational purposes. Adapt and modify for your specific needs.

1. Reference Only - This repository is for learning, not direct cloning
2. Placeholder Values - All credentials are examples only
3. Customize for Your Needs - Adapt patterns to your specific workflow
4. Security First - Always review and adjust permissions for your environment

1. Start Simple - Begin with settings.json permissions
2. Add Commands - Implement one custom command
3. Integrate Agents - Add relevant specialized agents
4. Automate - Set up hooks for your workflow
5. Iterate - Refine based on your experience

Built with ❤️ for the Claude Code community

Transform your AI-assisted development with production-grade configurations