KinOS v6 - AI Team Orchestration System

🌟 Overview

KinOS is an advanced AI team orchestration system that enables autonomous collaboration between specialized AI agents. It uses a file-based architecture to coordinate multiple GPT-4 agents working together on complex projects.

See it in action: https://nlr.ai/

✨ Key Features

• 🤖 Autonomous Agent Teams: Pre-configured specialized teams for different project types
• 📁 Directory-Based Operation: Uses current directory as mission context
• 🔄 Dynamic Resource Management: Automatic scaling and resource allocation
• 🔍 Intelligent Content Management: Built-in deduplication and content organization
• 🔗 Git Integration: Automatic version control and change tracking
• 📊 Progress Monitoring: Real-time status tracking and logging

📊 Project Structure

💡 Best Practices

Mission Definition

• Be specific and detailed about expected outputs and deliverables
• Include clear format and structure requirements
• Define validation criteria and constraints

Repository Preparation

• Add relevant reference materials and examples as text files
• Structure directories to match expected output
• Include any required templates or configurations

Interactive Guidance

• Use kin interactive to open chat sessions with the project
• Guide work in progress and clarify requirements
• Review and refine implementations

🚀 Getting Started

Prerequisites

• Python 3.8+
• OpenAI API key
• Perplexity API key (for research capabilities)
• Git installed
• Node.js and npm installed (for repository visualization)

Optional Model Providers

• Ollama: For local model execution

  • Install from Ollama.ai
  • Usage: --model ollama_chat/<model_name>

• OpenRouter: For additional model providers

  • Get API key from OpenRouter.ai
  • Usage: --model openrouter/<provider>/<model_name>

• Default: OpenAI GPT-4o-mini

  • Requires OpenAI API key
  • Most stable and tested option
  • Usage: --model gpt-4o-mini (default)

• Cairo graphics library (for SVG to PNG conversion):

  • Windows: Install GTK3 runtime from GTK for Windows
  • Linux: sudo apt-get install libcairo2-dev pkg-config python3-dev
  • macOS: brew install cairo pkg-config

Installation Steps

1. Verify Prerequisites:

   • Python 3.8+ installed (python --version)
   • Git installed (git --version)
   • Node.js and npm installed (node --version, npm --version)
   • Cairo graphics library installed:
     • Windows: Install GTK3 runtime from GTK for Windows
     • Linux: sudo apt-get install libcairo2-dev pkg-config python3-dev
     • macOS: brew install cairo pkg-config

2. Clone the Repository:

   git clone --recursive https://github.com/DigitalKin-ai/kinos.git
   cd kinos

3. Configure Environment:

   cp .env.example .env
   # Edit .env and add your API keys:
   # - OPENAI_API_KEY
   # - PERPLEXITY_API_KEY

4. Run Installation Script:

   • Windows: install.bat
   • Linux/Mac:

     chmod +x install.sh
     ./install.sh

The installation script will:

• Update git submodules
• Install Python dependencies
• Set up custom Aider
• Build repo-visualizer
• Add KinOS to your PATH

Starting Your First Project

1. Create a new project directory:

mkdir my_project
cd my_project

2. Create a aider.mission.md file:

# Create .aider.mission.md file
# This file describes what you want to accomplish
# Example content:
"""
# Project Mission: Create a Python Web Scraper

## Objective
Build a web scraper that can:
- Extract data from e-commerce websites
- Save results to CSV files
- Handle pagination and rate limiting
- Respect robots.txt

## Requirements
- Use Python with BeautifulSoup
- Include error handling
- Add comprehensive documentation
- Create unit tests
"""

3. Launch KinOS:

# Generate and start 6 parallel agents
kin run agents --generate --count 6

4. Monitor Progress:

• Check suivi.md for detailed logs
• View diagram.svg for your repository diagram
• Review todolist.md for pending tasks

📖 Usage

Basic Commands

# Launch with default configuration
kin run agents

# Launch with specific model
kin run agents --model gpt-4o-mini  # Default model
kin run agents --model gpt-4o  # Other models may be supported in future

# Use with local models via Ollama
kin run agents --model ollama_chat/llama3.1:70B  # Use local Llama 3.1

# Use with model routers
kin run agents --model openrouter/anthropic/claude-3-5-haiku  # Use Haiku 3.5 via OpenRouter

# Generate new agents
kin generate agents

# Launch an interactive session with the project
kin interactive

Required Environment Variables

OPENAI_API_KEY=your_openai_key_here        # Required for all operations
PERPLEXITY_API_KEY=your_perplexity_key_here # Required for research capabilities

Common Operations

1. Start a New Project

# Create project directory
mkdir my_project
cd my_project

# Create mission file
echo "Project mission details..." > .aider.mission.md

# Launch KinOS
kin run agents

# add --verbose to any command to get more info

🛠️ Core Components

Agent Types

1. SpecificationAgent: Requirements analysis and documentation
2. ManagementAgent: Project coordination and resource allocation
3. writingAgent: Content creation and documentation
4. EvaluationAgent: Quality assurance and testing
5. DeduplicationAgent: Content organization and redundancy management
6. chroniclerAgent: Progress tracking and logging
7. redundancyAgent: Backup and consistency management
8. ProductionAgent: Code/content generation
9. researcherAgent: Research and analysis
10. IntegrationAgent: System integration and deployment

Possible Use-cases

• 📚 Book Writing
• 🔬 Literature Reviews
• 💻 Coding
• Business Plans
• Company Documentation

🤝 Contributing

We welcome contributions! Feel free to reach out directly to me: contact on https://nlr.ai/

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Aider for enabling AI-assisted development
• The AutonomousAI community for pioneering autonomous AI development
• Claude for being an awesome collaborator

📞 Support

• Telegram: https://t.me/+KfdkWFNZoONjMTE0
• Website: https://nlr.ai/
• Patreon: https://www.patreon.com/c/kins_autonomousais/membership

🔮 Future Plans

• Packaged version
• GUI
• Improved agent convergence

---

Made with ❤️ by NLR, Claude & the KinOS Community