🚀 MARIA v3.3.0

Graph RAG Intelligence & Enhanced /init Commands - Revolutionary AI development platform with Graph RAG technology, intelligent codebase analysis, and dynamic guideline generation. Features advanced AST parsing, knowledge graphs, and state-of-the-art incremental learning capabilities.

🌐 Homepage: https://bonginkan.ai/

🧠 Revolutionary Graph RAG Technology

MARIA v3.3.0 introduces cutting-edge Graph RAG (Retrieval Augmented Generation) capabilities that transform how AI understands and works with your codebase:

┌─────────────────────────────────────────────────────────────┐
│  🔍 AST PARSING    →   📊 KNOWLEDGE GRAPH   →   🤖 AI GUIDE  │
│                                                             │
│  TypeScript/JS     →   Neo4j Relationships  →   Dynamic     │
│  Code Analysis     →   Semantic Connections →   Guidelines  │
│  Dependencies      →   Vector Embeddings    →   Smart Recs  │
└─────────────────────────────────────────────────────────────┘

🎯 What's New in v3.3.0

• 🧪 Graph RAG Engine: Deep codebase understanding through knowledge graphs
• 🔄 Smart /init Command: Complete project analysis with AST parsing
• 📈 /update Command: Intelligent incremental updates with delta detection
• 🎨 Enhanced CLI UX: Progress bars, spinners, and TTY/JSON output modes
• 🔍 Provenance Tracking: Full transparency in AI recommendations
• ⚡ State Persistence: Efficient change tracking with SHA-256 hashing

🎨 Beautiful CLI Experience

When you run maria, you'll see:

╔══════════════════════════════════════════════════════════╗
║  ███╗   ███╗ █████╗ ██████╗ ██╗ █████╗                  ║
║  ████╗ ████║██╔══██╗██╔══██╗██║██╔══██╗                 ║
║  ██╔████╔██║███████║██████╔╝██║███████║                 ║
║  ██║╚██╔╝██║██╔══██║██╔══██╗██║██╔══██║                 ║
║  ██║ ╚═╝ ██║██║  ██║██║  ██║██║██║  ██║                 ║
║  ╚═╝     ╚═╝╚═╝  ╚═╝╚═╝  ╚═╝╚═╝╚═╝  ╚═╝                 ║
║        AI-Powered Development Platform                   ║
║         (c) 2025 Bonginkan Inc.                          ║
╚══════════════════════════════════════════════════════════╝

MARIA CODE v3.3.0 — Graph RAG Ready
/help for commands | Providers: 8/8 OK | Graph RAG: Enabled

Available AI Providers:
☁️  Cloud AI: OpenAI, Anthropic, Google, Groq, xAI
💻 Local AI: Ollama, LM Studio, vLLM
🧠 Graph RAG: Knowledge Graphs, Vector Search, AST Analysis

⚡ Installation

# Global installation (recommended)
npm install -g @bonginkan/maria

# Or with pnpm
pnpm add -g @bonginkan/maria

🧬 Graph RAG Quick Start

Initialize Your Project with Graph RAG

# Full codebase analysis with knowledge graph construction
maria
/init

# Output:
# 🧠 Initializing Graph RAG analysis...
# 📁 Scanning project files...
# 🔍 Parsing AST structures...
# 📊 Building knowledge graph...
# ✨ Generating MARIA.md guidelines...

Incremental Updates

# Smart incremental updates with delta detection
/update

# With specific change detection
/update --since git:HEAD~3
/update --since 2025-08-20
/update --since state

🎯 Core Features

🧠 Graph RAG Intelligence

• AST Parsing: Deep TypeScript/JavaScript code analysis
• Knowledge Graphs: Neo4j-powered semantic relationships
• Vector Search: Hybrid search with RRF ranking
• Delta Detection: Three methods (git, time, state-based)
• Provenance: Full explanation of AI recommendations

🤖 AI Provider Support

8 AI Providers with automatic failover:

🎨 Advanced CLI Features

# Enhanced slash commands with Graph RAG
/init          # Full project analysis with knowledge graphs
/update        # Incremental updates with delta detection
/memory        # View Graph RAG memory statistics
/status        # System health with Graph RAG metrics
/help          # Enhanced help with visual diagrams

📊 Graph RAG Visualization

Knowledge Graph Structure:
├── 📁 Project Root
│   ├── 🔗 Dependencies
│   │   ├── External Packages
│   │   └── Internal Modules  
│   ├── 🧩 Components
│   │   ├── React Components
│   │   └── Utility Functions
│   ├── 🎯 Types & Interfaces
│   │   ├── API Contracts
│   │   └── Domain Models
│   └── 🔄 Data Flow
│       ├── State Management
│       └── Event Handlers

🚀 Advanced Usage

Custom Delta Detection

# Git-based changes
/update --since git:HEAD~5

# Time-based changes  
/update --since 2025-08-20T10:00:00Z

# State-based changes (SHA-256)
/update --since state

Output Modes

# TTY mode (interactive)
maria

# JSON mode (CI/CD)
MARIA_OUTPUT=json maria

Budget Control

// Budget-aware processing
{
  "budget": {
    "maxFiles": 1000,
    "maxTimeMs": 300000,
    "maxTokens": 100000
  }
}

🏗️ Architecture

Graph RAG Pipeline

Input → AST Parser → Knowledge Graph → Vector Store → AI Analysis → Guidelines
  ↓         ↓             ↓               ↓            ↓           ↓
Files → Structures → Relationships → Embeddings → Insights → MARIA.md

Memory Systems

• System 1: Fast pattern recognition, code snippets
• System 2: Deep reasoning, architectural analysis
• Graph Store: Persistent knowledge relationships
• Vector Store: Semantic similarity search

Safety & Privacy

• Local Processing: AST parsing runs locally
• Gradual Degradation: Works without external services
• Budget Controls: Prevents resource exhaustion
• Safe Defaults: Non-destructive operations

📚 Graph RAG Configuration

Neo4j Integration

{
  "graphRAG": {
    "neo4j": {
      "uri": "bolt://localhost:7687",
      "user": "neo4j",
      "password": "your-password"
    },
    "vectorDB": {
      "provider": "qdrant",
      "collection": "maria-knowledge"
    }
  }
}

AST Analysis Settings

{
  "astParser": {
    "languages": ["typescript", "javascript"],
    "includePatterns": ["src/**/*.ts", "src/**/*.tsx"],
    "excludePatterns": ["node_modules/**", "dist/**"],
    "maxDepth": 10
  }
}

🎯 Use Cases

🏢 Enterprise Development

• Code Review: AI-powered analysis with Graph RAG insights
• Architecture Planning: Relationship mapping and dependency analysis
• Knowledge Transfer: Dynamic guideline generation for teams
• Legacy Migration: Deep codebase understanding for modernization

👨‍💻 Individual Developers

• Project Onboarding: Instant understanding of new codebases
• Code Quality: Intelligent suggestions based on project patterns
• Documentation: Auto-generated guidelines that evolve with code
• Learning: Explore code relationships through knowledge graphs

📖 API Reference

Core Services

import { 
  IntelligentRouterService,    // Graph RAG routing
  DualMemoryEngine,           // System 1 & 2 memory
  FileSystemService           // Safe file operations
} from '@bonginkan/maria';

Graph RAG Services

import {
  GraphRAGService,            // Main Graph RAG orchestrator
  ASTAnalyzer,               // Code structure analysis
  KnowledgeGraphService,     // Neo4j graph management
  DeltaDetectorService       // Change detection
} from '@bonginkan/maria/graph-rag';

🚀 Changelog v3.3.0

✨ New Features

• Graph RAG Engine: Complete implementation with Neo4j integration
• Enhanced /init Command: AST parsing + knowledge graph construction
• New /update Command: Intelligent incremental updates
• Delta Detection: Git/time/state-based change tracking
• Advanced CLI UX: Progress bars, spinners, TTY/JSON modes
• Provenance System: Full transparency in AI recommendations

🔧 Improvements

• Build System: Fixed 870+ TypeScript syntax errors
• Import Paths: Cleaned up all module resolution issues
• Dependencies: Added Neo4j driver and Babel traverse support
• Memory Efficiency: Optimized AST parsing and graph operations
• Error Handling: Enhanced error reporting with suggestions

📦 Dependencies

• Added: neo4j-driver for graph database operations
• Added: @babel/traverse for AST parsing capabilities
• Updated: All provider integrations with improved stability

🤝 Contributing

We welcome contributions to MARIA's Graph RAG capabilities!

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-graph-rag)
3. Commit your changes (git commit -m 'Add amazing Graph RAG feature')
4. Push to the branch (git push origin feature/amazing-graph-rag)
5. Open a Pull Request

📄 License

MIT License - see LICENSE file for details.

🌟 Support

• 📧 Email: support@bonginkan.ai
• 🐛 Issues: GitHub Issues
• 💬 Discussions: GitHub Discussions
• 📚 Documentation: docs.bonginkan.ai

MARIA v3.3.0 - Revolutionizing AI development with Graph RAG intelligence 🚀