Transform Claude Code from a simple assistant into a team of 7 specialized AI agents working autonomously on your projects.
You: "I need an e-shop"
Claude: [Master โ Task โ Search โ Coding โ Data โ Master]
Result: Complete e-commerce solution with autonomous profile switchingA revolutionary system that gives you 7 specialized AI agents that automatically collaborate:
- ๐ Master - Universal coordinator and main communication hub
- ๐ Task - Project management, planning & workflow orchestration
- ๐ Search - Advanced research & information gathering with Matrix Searchโข
- ๐ป Coding - Professional development, debugging & architecture
- ๐ Data - Analytics, insights & data science workflows
- โ๏ธ Copywriter - Marketing copy, conversion optimization & brand voice
- ๐จ Design - UX/UI design, accessibility & user experience
Key Benefits:
- โ Autonomous operation - Agents switch automatically based on task needs
- โ Specialized expertise - Each agent is optimized for specific domains
- โ Seamless collaboration - Agents work together without user intervention
- โ Production-ready output - Professional-grade solutions from day one
- Claude Code installed (
npm install -g @anthropic-ai/claude-code) - Linux/Mac/WSL environment
-
Clone this repository
git clone https://github.com/kuzmany/claude-code-agents.git cd claude-code-agents -
Copy agent configurations to Claude directory
mkdir -p ~/.claude cp agents/* ~/.claude/
-
Start using
claude --dangerously-skip-permissions
The system is ready! Agents will switch automatically as needed.
The system operates through slash commands and autonomous agent switching:
/master # Switch to universal coordinator agent
/search # Switch to advanced research agent
/task # Switch to project management agent
/coding # Switch to professional development agent
/data # Switch to analytics & data science agent
/copywriter # Switch to marketing copywriter agent
/design # Switch to UX/UI design agent
/current # Show currently active agentSimply describe your need - the system automatically:
- Analyzes requirements
- Switches to appropriate expert agents
- Collaborates between agents as needed
- Delivers complete solutions
๐ง Matrix Searchโข - Advanced 3-level research system
- Level 1: Quick scan (0-3 seconds)
- Level 2: Deep dive with cross-referencing
- Level 3: Expert analysis with source verification
โก Autonomous Switching - No manual intervention needed
- Task analysis triggers appropriate agent changes
- Seamless handoffs between expert domains
- Context preservation across switches
๐ฅ Specialized Agents - Each optimized for specific workflows
- Search: Ultimate research tool with brutal honesty
- Task: Project orchestration and quality control
- Coding: Production-ready development practices
- Data: Advanced analytics and visualization
- Copywriter: Conversion-focused marketing copy and brand voice
- Design: User-centered design and accessibility expertise
All documentation is contained in this README. The system includes:
- 7 Specialized Agents - Each optimized for specific tasks
- Autonomous Switching - Agents change automatically based on context
- Matrix Searchโข - Advanced research capabilities with source verification
- Production Focus - Every output is ready for real-world implementation
- Slash Commands - Manual agent control when needed
User: "Analyze cryptocurrency market trends for 2025"
System: [Search] โ [Data] โ [Task] โ [Master]
Result: Comprehensive market analysis with predictionsUser: "Create a todo app with React and TypeScript"
System: [Task] โ [Coding] โ [Search] โ [Coding] โ [Master]
Result: Production-ready application with best practicesUser: "Help me plan a startup launch strategy"
System: [Search] โ [Task] โ [Data] โ [Master]
Result: Complete go-to-market strategy with data backingUser: "Create converting landing page copy for our SaaS product"
System: [Search] โ [Copywriter] โ [Design] โ [Task] โ [Master]
Result: High-converting landing page with optimized copy and UX designUser: "Design a user-friendly mobile app interface"
System: [Search] โ [Design] โ [Copywriter] โ [Coding] โ [Master]
Result: Complete UI/UX design with microcopy and implementation guideThe agents work together seamlessly:
User Request โ Master Analysis โ Expert Agent Selection โ Task Execution โ Results
โ โ โ โ โ
Autonomous Understands Switches to Specialized Complete
Operation Requirements Needed Expert Execution Solution
Agent Flow Examples:
- Complex Research: Search โ Data โ Task โ Master
- Software Project: Task โ Coding โ Search โ Coding โ Master
- Business Analysis: Search โ Data โ Task โ Coding โ Master
- Marketing Campaign: Search โ Copywriter โ Design โ Task โ Master
- Product Launch: Task โ Design โ Copywriter โ Data โ Master
"Brutal honesty over polite uncertainty. Real solutions over theoretical perfection."
System Principles:
- โ Autonomous operation - Agents switch without asking permission
- โ Specialized expertise - Each agent masters one domain
- โ Collaborative intelligence - Agents work together automatically
- โ Production focus - Every output is ready for real-world use
# 1. Clone repository
git clone https://github.com/kuzmany/claude-code-agents.git
cd claude-code-agents
# 2. Install agent configurations
mkdir -p ~/.claude
cp agents/* ~/.claude/
# 3. Start Claude Code
claude --dangerously-skip-permissions- Contributing: See CONTRIBUTING.md for guidelines
- License: MIT License - see LICENSE file for details
Created with โค๏ธ for the Claude Code community
โญ Star this repo if it saves you time!