/icon-agents

a collection of claude code AI agents and commands that embody legendary icons

Primary LanguageShellOtherNOASSERTION

Icon Agents 🏗️

Legendary programming wisdom at your fingertips

Transform your development workflow with 64 legendary experts across 8 domains, featuring intelligent multi-domain analysis and direct LLM expert selection.

🚀 Quick Start

Important: Install Icon Agents in your Claude Code project directory (where you run claude commands).

Installation (Choose One)

# Clone and install (recommended for developers)
git clone https://github.com/Commands-com/icon-agents.git
cd icon-agents && ./install.sh

# NPX (if you have Node.js)
npx icon-agents@latest

# One-line curl (works everywhere)
curl -fsSL https://raw.githubusercontent.com/Commands-com/icon-agents/refs/heads/main/install.sh | bash

# Manual download
wget https://raw.githubusercontent.com/Commands-com/icon-agents/refs/heads/main/install.sh
chmod +x install.sh && ./install.sh

🎯 What You Get

Main Command: Multi-Domain Analysis

  • /icon-review - Intelligent analysis that automatically selects experts across all 8 domains

Pod-Specific Commands (8 Specialized Reviews)

  • /icon-programming-review - Code analysis with 8 legendary programmers
  • /icon-security-review - Security analysis with 8 security experts
  • /icon-design-review - Design analysis with 8 design masters
  • /icon-business-review - Strategy analysis with 8 business leaders
  • /icon-data-ai-review - AI/ML analysis with 8 data science pioneers
  • /icon-product-policy-review - Product analysis with 8 policy experts
  • /icon-platform-operations-review - Infrastructure analysis with 8 platform architects
  • /icon-healthcare-review - Healthcare AI business analysis with 8 healthcare luminaries

Individual Expert Agents (64 Total)

All experts are available as individual Task agents for direct consultation.

🏗️ Architecture: 8 Pods of Excellence

Programming Pod (8 Experts)

  • Linus Torvalds - Engineering taste, systems programming, pragmatic design
  • John Carmack - Performance optimization, empirical analysis, low-level efficiency
  • Rich Hickey - Simplicity, complexity elimination, functional design
  • Alan Kay - System evolution, long-term vision, human-centered design
  • Kent Beck - Test-driven development, refactoring, feedback loops
  • Barbara Liskov - Data abstraction, contract design, type safety
  • Leslie Lamport - Distributed systems, formal correctness, concurrency
  • Donald Knuth - Algorithmic analysis, mathematical rigor, complexity theory

Security Pod (8 Experts)

  • Dan Kaminsky - DNS security, infrastructure vulnerabilities
  • Katie Moussouris - Vulnerability management, bug bounty programs
  • Bruce Schneier - Cryptography, privacy, security economics
  • Mikko Hyppönen - Malware analysis, threat intelligence
  • Tarah Wheeler - Cybersecurity policy, international cyber norms
  • Mudge Zatko - Security research, vulnerability discovery
  • Eva Galperin - Digital rights, stalkerware research
  • Moxie Marlinspike - Applied cryptography, Signal protocol

Design Pod (8 Experts)

  • Dieter Rams - Minimalist functionality, timeless aesthetics
  • Don Norman - Human-centered design, usability principles
  • Edward Tufte - Information design, data visualization
  • Jonathan Ive - Industrial design, material innovation
  • Susan Kare - Icon design, visual communication
  • Jakob Nielsen - Usability engineering, interface optimization
  • Kat Holmes - Inclusive design, accessibility excellence
  • Lou Downe - Service design, end-to-end user experiences

Business Pod (8 Experts)

  • Clayton Christensen - Disruptive innovation theory
  • Michael Porter - Competitive strategy, Five Forces analysis
  • Eric Ries - Lean startup methodology
  • Steve Jobs - Product vision, design excellence
  • Jeff Bezos - Customer obsession, long-term thinking
  • Satya Nadella - Transformation leadership, growth mindset
  • Reid Hoffman - Network effects, platform strategy
  • Elon Musk - First principles thinking, manufacturing innovation

Data & AI Pod (8 Experts)

  • Andrew Ng - Practical AI education, machine learning accessibility
  • Fei-Fei Li - Computer vision, human-centered AI
  • Geoffrey Hinton - Deep learning principles, neural networks
  • Hilary Mason - Data strategy, business applications
  • Yann LeCun - Convolutional networks, computer vision
  • Cassie Kozyrkov - Decision science, statistical thinking
  • DJ Patil - Data science leadership, social impact
  • Demis Hassabis - AGI research, neuroscience-inspired AI

Product & Policy Pod (8 Experts)

  • Marty Cagan - Product management, continuous discovery
  • Gene Kim - DevOps, systems thinking
  • Joanna Bryson - AI governance, algorithmic accountability
  • Michelle Zatlyn - Scaling expertise, operational excellence
  • Julie Zhuo - Design leadership, user experience strategy
  • Ben Horowitz - Crisis leadership, organizational culture
  • Tristan Harris - Humane technology, attention economy ethics
  • Cathy O'Neil - Algorithmic accountability, bias detection

Platform & Operations Pod (8 Experts)

  • Tim Berners-Lee - Web architecture, open standards
  • Vint Cerf - Internet protocols, global networking
  • Radia Perlman - Network engineering, fault tolerance
  • Werner Vogels - Cloud architecture, distributed systems
  • Martin Fowler - Enterprise architecture, evolutionary design
  • Brendan Gregg - Performance engineering, systems observability
  • Kelsey Hightower - Cloud native, Kubernetes mastery
  • Jessie Frazelle - Systems security, container security

Healthcare Pod (8 Experts)

  • Atul Gawande - Healthcare systems optimization, patient safety, clinical implementation
  • Eric Topol - Digital medicine, AI in healthcare, personalized medicine
  • Regina Barzilay - AI for drug discovery, medical imaging AI, clinical decision support
  • Daphne Koller - Precision medicine, computational biology, AI-driven drug discovery
  • Bob Wachter - Digital health transformation, clinical informatics, healthcare IT
  • Fei-Fei Li - Medical imaging AI, ambient intelligence in healthcare, ethical AI
  • Andrew Ng - Healthcare AI deployment, scalable medical AI, clinical implementation
  • Vinod Khosla - Healthcare venture capital, digital health business models, innovation scaling

🧠 How It Works

Intelligent Expert Selection

No more choosing between experts or running all of them. The system:

  1. Analyzes your problem using direct LLM intelligence (no fragile keyword matching)
  2. Selects 2-3 most relevant experts from the appropriate domain(s)
  3. Executes in parallel for fast, comprehensive analysis
  4. Synthesizes results into unified, actionable recommendations

Example Analysis Flow

Your Input: "Slow database queries in our microservices architecture"

AI Analysis: 
- Primary domain: Performance + Architecture + Platform
- Selected experts: John Carmack (performance), Rich Hickey (architecture), Werner Vogels (distributed systems)

Parallel Execution:
🚀 Carmack analyzes query optimization and caching strategies
🧠 Hickey examines architectural complexity and data flow
☁️ Vogels reviews distributed system patterns and consistency

Unified Output: Concrete recommendations with priority levels

📋 Usage Examples

Multi-Domain Analysis

/icon-review
# Analyzes your project and automatically selects relevant experts across all domains

Domain-Specific Analysis

/icon-programming-review --source ./src/performance-critical.js
# Focus on programming expertise for code analysis

/icon-security-review --source ./auth-module/
# Security-focused analysis of authentication code

/icon-design-review --source ./mockups/user-flow.figma
# Design analysis of user experience flows

Individual Expert Consultation

# Direct expert access via Task agents
Task(linus-torvalds): "Review this data structure design"
Task(john-carmack): "Analyze these performance bottlenecks"
Task(rich-hickey): "Evaluate this architecture for simplicity"

🎛️ Installation Options

The interactive installer will ask you:

  • Installation directory (current directory by default)
  • Which pods to install (all 8 pods or specific ones)
  • Confirmation before proceeding

Pod Selection

  • all - Install all 8 pods (64 experts)
  • Individual pods - Choose specific domains:
    • programming - 8 legendary programmers
    • security - 8 security experts
    • design - 8 design masters
    • business - 8 business strategists
    • data-ai - 8 AI/ML pioneers
    • product-policy - 8 product/policy experts
    • platform-operations - 8 platform architects
    • healthcare - 8 healthcare luminaries

🚀 After Installation

1. Restart Claude Code

# Restart to load new agents and commands

2. Start Getting Legendary Wisdom

# Multi-domain analysis
/icon-review

# Domain-specific analysis  
/icon-programming-review
/icon-security-review

🗑️ Uninstalling

To remove Icon Agents from your project:

# NPX (recommended)
npx icon-agents uninstall

# Run the uninstall script directly
./uninstall.sh

# Or if downloaded separately
curl -fsSL https://raw.githubusercontent.com/Commands-com/icon-agents/refs/heads/main/uninstall.sh | bash

The uninstall script will:

  • Remove all Icon Agents expert files and commands
  • Offer to clean up empty directories
  • Preserve any other Claude Code configurations

🎯 Key Features

  • ✅ No Keyword Matching - Direct LLM expert selection (no fragile patterns)
  • ✅ Parallel Execution - Multiple experts analyze simultaneously (3x faster)
  • ✅ Intelligent Selection - 2-3 most relevant experts, not all 56
  • ✅ Manual Overrides - Force specific experts when needed
  • ✅ Multi-Domain Coverage - 7 specialized domains of expertise
  • ✅ Unified Recommendations - Synthesized insights with priority levels
  • ✅ Individual Access - All 56 experts available as Task agents

🏆 Why Icon Agents?

Before: Tool Proliferation

  • Separate tools for code review, security analysis, design feedback
  • Manual expert selection and sequential execution
  • Disconnected insights and conflicting recommendations

After: Unified Legendary Wisdom

  • One system that understands your multi-domain problems
  • Intelligent orchestration of the right experts at the right time
  • Parallel analysis with synthesized, actionable recommendations
  • Legendary expertise from the most influential minds in technology

📚 Documentation

  • Commands: Detailed command documentation in commands/*/README.md
  • Agents: Individual expert capabilities in agents/*/[expert].md
  • Architecture: Pod-based design principles in docs/ARCHITECTURE.md

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines on:

  • Adding new legendary experts
  • Improving expert selection algorithms
  • Enhancing domain coverage
  • Optimizing parallel execution

📄 License

MIT License - See LICENSE for details.

🌟 Support the Project

If Icon Agents helps your development workflow, consider:

  • ⭐ Starring the repository
  • 🐛 Reporting issues and suggestions
  • 🔄 Sharing with fellow developers
  • 💡 Contributing new expert domains

Transform your development workflow with legendary wisdom. Install Icon Agents today and code with the insights of the greatest minds in technology.

npx icon-agents@latest

Legendary expertise, intelligently orchestrated. 🏗️