A multi-model AI coding assistant CLI with MCP (Model Context Protocol) support. Inspired by Claude Code, Koli Code supports multiple AI providers including OpenAI, Anthropic, Google Gemini, and xAI's Grok.

• 🤖 Multi-Model Support: Switch between OpenAI GPT-4, Anthropic Claude, Google Gemini, and xAI Grok
• 🔌 MCP Protocol: Connect to external tools and data sources via Model Context Protocol
• ⚡ Streaming Responses: Real-time streaming for all supported models
• 🛡️ Security First: Built-in security rules to prevent malicious code generation
• 📝 Concise Responses: Optimized for terminal use with minimal output
• 🔧 Extensible: Easy to add new models and tools

npm install -g koli-code

1. Set up your API keys:

export OPENAI_API_KEY=your_openai_key
export ANTHROPIC_API_KEY=your_anthropic_key
export GOOGLE_API_KEY=your_google_key
export GROK_API_KEY=your_grok_key

2. Run Koli Code:

koli

Create a .env file in your project root:

# Model API Keys
OPENAI_API_KEY=your_openai_api_key_here
ANTHROPIC_API_KEY=your_anthropic_api_key_here
GOOGLE_API_KEY=your_google_api_key_here
GROK_API_KEY=your_grok_api_key_here

# Default Model (openai, anthropic, gemini, grok)
DEFAULT_MODEL=grok

# Clone the repository
git clone https://github.com/yourusername/koli-code.git
cd koli-code

# Install dependencies
npm install

# Build the project
npm run build

# Run in development mode
npm run dev

• Model Abstraction Layer: Unified interface for all AI models
• Prompt Engineering: Optimized prompts for coding tasks
• MCP Support: Connect to databases, APIs, and external tools
• Tool System: Modular tools for file operations, code execution, and more

MIT