Claude Code AI Agent System

A comprehensive ecosystem of 71 specialized AI agents for autonomous software development with Claude Code. Each agent is an expert in their domain, working individually or in coordination to deliver production-ready solutions.

📁 Repository Structure

ClaudeAgents/
├── agents/              # 71 specialized agent definitions
├── commands/            # Organized workflow commands
│   ├── development/     # Code review, debugging, refactoring
│   ├── quality/         # Testing, security, performance audits
│   ├── deployment/      # Deploy prep, infrastructure setup
│   ├── seo/             # SEO auditing and optimization
│   ├── specialized/     # Language-specific tools (Rust, Python, etc.)
│   └── workflows/       # Multi-agent orchestration patterns
├── tools/               # Validation and utilities
│   └── validate_agents.py  # Agent consistency validation
├── examples/            # Design specifications and prototypes
│   └── analysis/        # Project analysis system (prototype)
├── docs/                # Comprehensive documentation
│   ├── manifesto.md     # Professional principles
│   ├── architecture.md  # System design and patterns
│   └── contributing.md  # Contribution guidelines
├── CLAUDE.md           # Core project instructions for Claude Code
└── TODO.md             # Improvement roadmap and tasks

🚀 Quick Start

New to ClaudeAgents? Get started in 3 minutes:

👉 Read the Quick Start Guide - Your first workflow in <5 minutes

Find a Workflow:

• 📋 Command Catalog - Browse all 59 workflows by category
• 🔍 Search by problem - “I need to build an API” → /api-design
• 🎯 Search by technology - “React, PostgreSQL” → relevant workflows

Common Scenarios:

• Building a web app? → /saas-mvp (8-12 hours, full-stack)
• Need an API? → /api-design (4-6 hours, REST/GraphQL)
• Testing strategy? → /quality:testing-strategy (4-8 hours, comprehensive)
• Performance issues? → /quality:performance-optimization (2-3 hours audit)
• Security audit? → /security-audit (4-6 hours, vulnerability assessment)

💡 Pro Tip: Start with a workflow (commands/), then explore individual agents for custom orchestration.

📚 Documentation

Getting Started:

• 🚀 Quick Start Guide - Your first workflow in 3 minutes
• 📋 Command Catalog - All 59 workflows searchable by category
• 🎯 Workflow Examples - Real case studies with metrics (coming soon)

Core Documentation:

• The Manifesto - Professional principles and engineering standards
• System Architecture - Design patterns, component interactions
• Users’ Guide - Comprehensive guide for using agents
• Contributing Guide - How to add agents and improvements

Developer Resources:

• Agent Tiers - Quality-based 3-tier organization
• Telemetry Guide - Privacy-first usage tracking
• Development Process - Contributor workflow

Privacy & Trust:

• Telemetry Privacy - What we collect, why, and how to opt-in

🎯 Agent Categories

📊 Strategy & Planning

• product-strategist - Market research, competitive analysis, product ideation
• project-orchestrator - Complex project coordination, multi-agent orchestration

🏗️ Core Development

• full-stack-architect - Web applications (React, Next.js, Svelte + backends)
• backend-api-engineer - Backend APIs (REST, GraphQL, microservices, server-side logic)
• mobile-developer - iOS/Android (Swift, Kotlin, React Native, Flutter)
• game-development-engineer - Unity/Unreal Engine, 2D/3D games, mobile/VR/AR, multiplayer

🏭 Infrastructure & Platform

• cloud-architect - Multi-cloud strategy, AWS/Azure/GCP architecture, cost optimization
• devops-engineer - CI/CD, Docker, Kubernetes, cloud deployment
• infrastructure-as-code-specialist - Terraform, Pulumi, multi-cloud IaC, state management, policy as code
• platform-engineering-specialist - Internal Developer Platforms (IDPs), Backstage, golden paths, developer experience
• edge-computing-specialist - Cloudflare Workers, Deno Deploy, Vercel Edge, <50ms global latency
• linux-sysadmin - System administration, OS hardening, server configuration
• macos-specialist - macOS administration, MDM integration, Homebrew, Apple Silicon optimization, enterprise Mac fleets
• windows-specialist - Windows Server, Active Directory, Group Policy, PowerShell DSC, enterprise Windows infrastructure

💾 Database & Streaming

• data-engineer - Data pipelines, analytics, ML infrastructure
• database-administrator - Production database operations, performance tuning, backup/recovery
• postgresql-expert - Advanced PostgreSQL optimization, query tuning, replication, high availability, pgvector/PostGIS/TimescaleDB
• kafka-expert - Apache Kafka architecture, stream processing, event-driven systems, Kafka Streams/Connect, real-time pipelines

🤖 AI & Machine Learning (Pattern-Based Specialists)

• llm-integration-architect - Multi-model LLM integration, intelligent routing, cost optimization, vendor-neutral architecture
• prompt-engineering-specialist - Advanced prompting techniques, chain-of-thought, few-shot learning, prompt optimization across all LLMs
• rag-systems-engineer - Retrieval-Augmented Generation architecture, vector databases, hybrid search, document processing pipelines
• fine-tuning-specialist - LLM fine-tuning (LoRA, QLoRA), dataset engineering, RLHF/DPO, custom model deployment
• inference-optimization-specialist - Self-hosted LLM deployment, quantization, GPU optimization, vLLM, TensorRT-LLM, cost-efficient serving
• generative-image-specialist - Image generation (Stable Diffusion, DALL-E, Midjourney), LoRA training, ControlNet, ComfyUI workflows
• systems-engineer - Rust, C++, Go, performance-critical systems
• metaprogramming-specialist - Lisp, macros, DSLs, code generation

₿ Blockchain & Web3

• blockchain-web3-engineer - Solidity smart contracts, Solana/Rust, DeFi, NFTs, dApp development, Web3 security

🚨 Operational Excellence

• debugging-specialist - Advanced debugging across all languages, root cause analysis, memory/performance profiling
• observability-engineer - Full-stack observability, distributed tracing, SLO/SLI engineering, error budgets
• incident-coordinator - Incident response, war room coordination, postmortems, on-call engineering
• site-reliability-engineer - SRE methodology, error budgets, toil reduction, capacity planning, operational maturity
• developer-experience-engineer - API/SDK usability, developer onboarding, error message design, tooling ergonomics, DX metrics

🔒 Quality & Security

• security-audit-specialist - Vulnerability assessment, compliance
• dependency-security-specialist - Supply chain security, SBOM generation (CycloneDX/SPDX), license compliance, SLSA framework
• compliance-automation-engineer - SOC 2, HIPAA, PCI-DSS, GDPR automation, continuous compliance, audit readiness
• qa-test-engineer - Testing strategies, automation, quality assurance
• test-automation-engineer - Playwright/Cypress E2E, visual regression, API testing, CI/CD integration
• accessibility-expert - WCAG compliance, inclusive design
• frontend-performance-specialist - Core Web Vitals, bundle optimization, rendering performance
• code-architect - Architecture review, code quality, maintainability

🔍 SEO & Optimization

• seo-meta-optimizer - Meta tags, Open Graph, structured data, CTR optimization
• seo-technical-auditor - Crawlability, indexability, sitemaps, mobile-friendliness
• seo-performance-specialist - Core Web Vitals for rankings, TTFB, mobile-first performance
• seo-keyword-strategist - Keyword research, search intent analysis, competitive keyword gaps
• seo-content-optimizer - On-page optimization, readability, E-E-A-T, featured snippets
• seo-structure-architect - Site architecture, internal linking, URL structure, content silos

📋 Business Operations

• business-analyst - Requirements gathering, stakeholder management, BRD, user stories
• technical-writer - API docs, user guides, tutorials, developer documentation
• product-manager - Product roadmap, feature prioritization, OKRs, product metrics

🎨 Creative & Specialized

• digital-artist - UI/UX graphics, game assets, visual design
• video-director - Video production, editing, post-production
• audio-engineer - Audio production, Logic Pro, CoreAudio
• 3d-modeler - 3D assets, Blender workflows, game development
• comedy-writer - Creative writing, humor, narrative structures
• tv-writer - Television scripts, procedural dramas

🔧 Specialized Development

• functional-programmer - Haskell, Clojure, F#, type systems
• legacy-specialist - Legacy code migration, compatibility
• platform-integrator - Native platform APIs (macOS, Windows, Linux)
• embedded-iot-developer - C/C++ firmware, ESP32/STM32/Arduino, RTOS, MQTT/BLE, power optimization
• elisp-specialist - Emacs configuration, package development
• merge-conflict-resolver - Git conflicts, code integration

⚖️ Decision Support

• the-critic - Technical decision analysis, architectural critique
• creative-catalyst - Creative problem-solving, lateral thinking
• the-skeptic - NEW (Phase 3) - Questions automation necessity, recommends alternatives, radical honesty about AI limitations

💡 Usage Examples

Web Application Development

"Use project-orchestrator to plan a task management app with React frontend and Node.js backend"

Mobile App Creation

"Have mobile-developer create a cross-platform fitness tracking app"

AI Feature Integration

"Get ai-ml-engineer to implement semantic search with RAG for my documentation"

Security Review

"Ask security-audit-specialist to review my authentication system"

Architecture Decision

"Use the-critic to evaluate PostgreSQL vs MongoDB for real-time analytics"

📋 Available Commands (59 total)

Development Commands (commands/development/)

• api-design - REST/GraphQL API design with OpenAPI specs
• database-design - Database schema design with migrations and indexing
• database-optimization - Database performance tuning and optimization
• debug-help - Debugging assistance
• refactor-component - Code refactoring
• documentation-generator - Auto-generate docs
• git-workflow - Git operations
• cross-paradigm-translator - Language translation
• tech-debt-impact-measurement - NEW (Tier 1 Innovation) - Empirical technical debt cost quantification with business impact metrics

Quality Commands (commands/quality/)

• code-review - Comprehensive code review (consolidated from 3 specialized review commands)
• testing-strategy - Complete testing strategy design with CI/CD integration (consolidated from 3 testing commands)
• performance-optimization - Complete performance optimization workflow (consolidated from 3 performance commands)
• security-audit - Security vulnerability scan
• production-readiness - Deployment checklist
• dependency-audit - Dependency security check
• infrastructure-audit - IaC and cloud infrastructure assessment (security, cost, reliability)
• cognitive-load-optimization - NEW (Tier 1 Innovation) - Industry-first cognitive complexity measurement and reduction strategies
• compliance-audit-soc2 - NEW - SOC 2 compliance preparation and audit readiness

Operations Commands (commands/operations/)

• monitoring-stack-setup - NEW - Complete observability implementation (Prometheus, Grafana, OpenTelemetry, distributed tracing)
• incident-response-workflow - NEW - Production incident management with war room coordination and postmortems
• disaster-recovery-plan - NEW - Business continuity and disaster recovery planning with RTO/RPO targets
• production-learning-loop - NEW (Tier 1 Innovation) - Self-improving organizational memory system with postmortem automation

Deployment Commands (commands/deployment/)

• deploy-prep - Deployment preparation
• dokku-deploy - Dokku deployment
• orb-stack - OrbStack configuration
• ssh-pi-ops - Raspberry Pi operations

Specialized Commands (commands/specialized/)

• Python tools (uv-workflow, modern-stack, data-pipeline, web-api, scraping)
• rust-cargo - Rust development
• xcode-power-tools - iOS development
• safari-web-extension - Safari extensions
• lisp-macro-workshop - Lisp macros
• roswell - Common Lisp setup

Workflow Commands (commands/workflows/)

• ai-agent-council - Multi-agent collaboration
• team-comm-hub - Team communication
• crisis-manager - Emergency response
• ai-code-battle - Code comparison
• microservices-architecture - Complete microservices design with service mesh
• platform-migration - Platform migration strategy and execution
• streaming-architecture - Real-time streaming architecture design
• debate - NEW (Phase 3) - Agent conflict theater for technical decisions (45-90 min, surfaces hidden tradeoffs)

SEO Commands (commands/seo/)

• comprehensive-seo-audit - Full site SEO health check (seo-technical-auditor, seo-meta-optimizer, seo-performance-specialist)
• content-optimization - Individual page optimization (seo-keyword-strategist, seo-content-optimizer, seo-meta-optimizer)
• keyword-research - Comprehensive keyword strategy (seo-keyword-strategist, seo-technical-auditor, seo-structure-architect)
• site-architecture-audit - Site structure optimization (seo-structure-architect, seo-technical-auditor, seo-keyword-strategist)

Business Commands (commands/business/)

• requirements-analysis - Complete BRD creation (business-analyst, product-manager, technical-writer)
• product-roadmap - Strategic product planning (product-manager, business-analyst, product-strategist)

Vertical Workflow Packages (commands/vertical/) NEW - Phase 2 & 3

• saas-mvp - Complete SaaS product development (8-12 hours, 6-8 agents, market strategy to production deployment)
• ecommerce-platform - E-commerce store launch (10-14 hours, 7-9 agents, mobile storefront to PCI compliance)
• fintech-compliance - NEW (Phase 3) - FinTech compliance & regulatory workflow (12-16 hours, 8-10 agents, PCI DSS/SOC 2/BSA-AML/GDPR/PSD2)

Platform Improvements (Recent Updates)

Command Consolidation for Coherence:

• Merged 12 redundant commands into 4 focused workflows (code-review, testing-strategy, performance-optimization, database-optimization)
• Removed overlapping database-review command (functionality integrated into database-optimization and infrastructure-audit)
• Result: Clearer command selection, reduced cognitive overhead, improved discoverability

Strategic Additions (7 New Commands):

• 3 Tier 1 Innovations: Industry-first capabilities (cognitive-load-optimization, production-learning-loop, tech-debt-impact-measurement)
• 4 High-Priority Operations: Production-ready operational excellence (monitoring-stack-setup, incident-response-workflow, disaster-recovery-plan, compliance-audit-soc2)
• Focus: Measurable business impact, empirical data-driven decision making, continuous improvement culture

🔍 Agent Selection Guide

The system uses intelligent agent selection based on:

• Keywords: Technology mentions (React, Python, database, etc.)
• Task Type: Development, testing, deployment, analysis
• Complexity: Single agent vs multi-agent orchestration
• Domain: Web, mobile, AI/ML, infrastructure, etc.

Agent selection is based on keyword analysis, task type, and domain expertise.

🛠️ Tools & Utilities

Installation

# Install required dependencies for validation tools
pip install -r tools/requirements.txt

Agent Validation

python3 tools/validate_agents.py

Validates all agent definitions for consistency and completeness.

Agent Registry (NEW - Phase 1)

Fast agent discovery through semantic indexing:

# View registry statistics
python3 tools/agent_registry.py stats

# Search for agents
python3 tools/agent_registry.py search "react mobile app"

# Find by capability/keyword/domain
python3 tools/agent_registry.py find optimization

Features:

• O(1) capability-based lookup (vs O(n) scanning)
• Semantic search with relevance scoring
• 1,290 keywords indexed across 50 agents
• Multi-index: capabilities, keywords, domains

Telemetry (Optional - Privacy-First, Enhanced Phase 3)

Opt-in usage tracking with comprehensive performance metrics:

# Enable telemetry (disabled by default)
python3 tools/telemetry.py enable

# View your usage statistics with performance metrics
python3 tools/telemetry.py summary

# Disable telemetry
python3 tools/telemetry.py disable

Privacy Promise:

• No PII, code snippets, or project details
• All data stored locally in ~/.claude-telemetry/
• Completely optional and transparent
• See Telemetry Guide for details

Performance Metrics (NEW):

• Percentile analysis (p50, p95, p99 duration)
• Fastest/slowest agent identification
• Performance trend tracking for optimization

Intelligent Orchestrator (NEW - Phase 2, Enhanced Phase 3)

Context-aware agent selection with tier-based prioritization:

# Auto-select agents based on project and request
python3 tools/intelligent_orchestrator.py "implement authentication with security best practices"

# Analyzes your project structure and selects optimal agents
python3 tools/intelligent_orchestrator.py "optimize frontend performance"

Features:

• Detects project type (web, mobile, data, ml, backend)
• Parses intent from natural language (implement, review, debug, optimize)
• Selects core + quality + support agents automatically
• Tier-based prioritization (Core > Extended > Experimental)
• Shows tier badges in output (⭐ CORE, ✓ EXTENDED, 🧪 EXPERIMENTAL)
• Generates workflow with reasoning and success criteria
• Estimates duration and defines prerequisites

Agent Emergence Tracking (NEW - Phase 3)

Organic agent evolution through usage pattern tracking:

# View emergence tracking dashboard
python3 tools/agent_emergence.py dashboard

# See patterns approaching promotion threshold
python3 tools/agent_emergence.py candidates

# Promote emergent agent to permanent status
python3 tools/agent_emergence.py promote <agent-name>

Features:

• Tracks agent usage gaps and combination patterns
• Auto-synthesizes composite agents when threshold met (10+ uses, 70%+ satisfaction, 5+ distinct use cases)
• Identifies unmet needs organically from real usage
• Prevents agent proliferation (only promotes validated patterns)
• Storage in ~/.claude-telemetry/emergence/

Analytics Dashboard (NEW - Phase 3)

Comprehensive visualization and analysis of agent usage patterns:

# View formatted analytics dashboard
python3 tools/analytics_dashboard.py

# Export analytics report as JSON
python3 tools/analytics_dashboard.py json

Features:

• Agent Rankings: Top agents by usage, success rate, and tier
• Tier Analysis: Distribution and performance by tier (Core, Extended, Experimental)
• Performance Trends: Fastest/slowest agents, outlier detection
• Quality Metrics: High performers (>95% success) and underperformers (<75%)
• Emergence Insights: Top composite patterns, promotion candidates
• Recommendations: Tier promotions/demotions, performance optimization targets
• Multi-format output: Human-readable dashboard or JSON export

Examples and Prototypes

The examples/ directory contains design specifications and proof-of-concept implementations:

• Project analysis system: Prototype for ML-based agent recommendations
• Implementation examples: Design patterns for future features

Note: These are not production-ready. See examples/README.md for details.

🎯 Model Assignment & Cost Optimization

ClaudeAgents uses strategic model assignment for cost optimization:

• 9 Haiku agents (18%) - Creative and specialized tasks
• 32 Sonnet agents (63%) - Development, coordination, and operational excellence
• 10 Opus agents (19%) - Complex analysis and critical systems (includes the-skeptic)

This distribution achieves ~75% cost savings vs using Opus for all tasks while maintaining appropriate capability levels.

See Model Assignment Strategy for details.

🏆 Success Metrics

Track agent effectiveness through:

• Feature Completion Rate: Delivered features working as specified
• Code Quality Scores: Maintainability, security, performance
• Integration Success: Seamless agent handoffs
• Development Velocity: Time from concept to production

Each agent is designed to deliver production-ready solutions with appropriate documentation, testing strategies, and deployment guidance for their domain.

---

For Claude Code-specific instructions and project context, see CLAUDE.md