Package Exports

@bonginkan/maria
@bonginkan/maria/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@bonginkan/maria) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

MARIA CODE - Advanced AGI Development Assistant

🤖 MARIA CODE is an advanced AGI (Artificial General Intelligence) development assistant that brings senior engineer-level capabilities to your terminal. With support for 14+ AI models, intelligent routing, and comprehensive development tools, MARIA transforms how you write, review, and deploy code.

🚀 Quick Start

# Install globally
npm install -g @bonginkan/maria@latest

# Initialize in your project
maria init

# Start interactive mode
maria chat

✨ Key Features

🧠 Interactive Router System

Natural Language Understanding: Interprets developer intentions and automatically routes to optimal commands
Context Awareness: Uses conversation history and project state for intelligent decision making
Multi-Step Execution: Decomposes complex tasks into executable steps
Auto Mode: Natural language → Automatic command execution

🤖 Multi-Model AI Support

Cloud Providers: OpenAI GPT-4o, Anthropic Claude, Google Gemini, Groq
Local Models: LM Studio, vLLM, Ollama (GPT-OSS, Qwen, Mistral)
Automatic Fallback: Seamlessly switches between providers
Task-Based Selection: Chooses optimal model for each task type

🎬 Media Generation

Video Generation: Text-to-Video and Image-to-Video with Wan 2.2
Image Generation: High-quality images with Qwen-Image
Batch Processing: Generate multiple variations
Resolution Control: 720p/1080p video, up to 1024x1024 images

⚡ Development Tools

Code Generation: Create entire features with context awareness
Code Review: Comprehensive analysis with actionable suggestions
Test Generation: Unit, integration, and E2E tests
Smart Commits: AI-generated conventional commit messages

📚 Commands

Core Commands

maria init              # Initialize MARIA in your project
maria chat              # Interactive mode (40+ slash commands)
maria code "prompt"     # Generate code with AI
maria vision image.png  # Analyze images with vision models
maria review            # Review code changes
maria test              # Generate tests
maria commit            # Create AI commit messages

Media Generation

maria video "A futuristic city"     # Generate AI video
maria image "Abstract art"         # Generate AI image

Interactive Slash Commands

In maria chat mode, use these commands:

/help - Show all commands
/status - System status
/model - Select AI model
/clear - Clear conversation
/config - Configuration panel
/video - Generate video
/image - Generate image
/pr-comments - Analyze PR comments
/review - Execute PR review
/exit - Exit interactive mode

🔧 Configuration

MARIA uses .maria-code.toml for configuration:

[project]
name = "my-project"
type = "typescript"

[ai]
default_provider = "openai"
fallback_provider = "local"

[ai.providers.openai]
api_key = "${OPENAI_API_KEY}"
model = "gpt-4o"

[ai.providers.local]
endpoint = "http://localhost:1234"
model = "gpt-oss-20b"

🌟 Example Usage

Create a REST API

maria code "Create a REST API with user authentication" --language typescript

Generate Tests

maria test src/ --framework jest --coverage

AI Video Generation

maria video "A red sports car racing through mountains" --model wan22-14b

Code Review

maria review --diff --suggestions --severity warning

📦 Installation Options

# Latest stable
npm install -g @bonginkan/maria@latest

# Development version
npm install -g @bonginkan/maria@alpha

# Beta version
npm install -g @bonginkan/maria@beta

# Run without installing
npx @bonginkan/maria@latest

🔌 AI Provider Setup

OpenAI

export OPENAI_API_KEY=your_key

Anthropic

export ANTHROPIC_API_KEY=your_key

Local Models (LM Studio)

Download LM Studio
Load a model (e.g., GPT-OSS, Mistral)
Start the server
MARIA auto-detects at http://localhost:1234

🎯 Use Cases

For Developers

Generate boilerplate code instantly
Review code for best practices
Create comprehensive test suites
Debug complex issues with AI assistance

For Teams

Standardize commit messages
Automate code reviews
Generate documentation
Create consistent APIs

For Learning

Understand new codebases quickly
Learn best practices
Get explanations for complex code
Practice with AI pair programming

🛠️ System Requirements

Node.js: 18.0.0 or higher
OS: macOS, Linux, Windows (WSL recommended)
Memory: 4GB RAM minimum (8GB+ for local models)
Storage: 500MB for CLI, varies for local models

🤝 Contributing

We welcome contributions! See Contributing Guide for details.

📄 License

MIT License - see LICENSE for details.

🔗 Links

💬 Support

Email: maria@bonginkan.ai
GitHub Issues: Create an issue