NextLovable — LangGraph Multi-Provider AI Agent

This project demonstrates a complete LangGraph-based multi-agent system with MongoDB persistence, ChromaDB vector storage, and support for multiple AI providers (Google Gemini and OpenRouter).

Features

• 🤖 Multi-Agent System: LangGraph workflow with planner → executor → reviewer → fixer auto-iteration
• 💾 Persistent Storage: MongoDB with Prisma for conversations and project data
• 🧠 Vector Memory: ChromaDB integration for long-term project context and code search
• 🔄 Auto-Iteration: Automatic retry and fix cycles when code generation fails
• 🎯 18 Advanced Tools: File operations, validation, optimization, and more
• 📱 Chat Interface: Modern UI with project management and session persistence
• 🐍 Python Backend Option: FastAPI backend with MCP (Model Context Protocol) integration

Quick Start

Option 1: Next.js Backend (Default)

1. Install dependencies:

   pnpm install

2. Set up environment variables:

   cp .env.example .env
   # Edit .env with your API keys

3. Start MongoDB (if using local instance):

   mongod

4. Run the application:

   pnpm run dev

Option 2: Python FastAPI Backend (Recommended)

1. Install Python dependencies:

   cd Pythonagents/fastapi-mcp-agent
   python3 -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt
   cd ../..

2. Set up environment variables:

   cp .env.example .env
   # Edit .env and set BACKEND_TYPE=python
   # Add your API keys (GROQ_API_KEY, GEMINI_API_KEY, etc.)

3. Run with Python backend:

   pnpm run dev:python

   Or manually:

   # Terminal 1: Start Python backend
   cd Pythonagents/fastapi-mcp-agent
   source venv/bin/activate
   uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
   
   # Terminal 2: Start Next.js frontend
   npm run dev

Docker Deployment

# Build and run both services
docker-compose up --build

# Or run individually
docker-compose up frontend
docker-compose up backend

API Providers

The system automatically selects the best available AI provider:

Primary: Google Gemini (Recommended)

• Environment Variable: GEMINI_API_KEY
• Get API Key: Visit Google AI Studio
• Advantages: No rate limiting, high quality, cost-effective
• Default Model: gemini-1.5-pro

Fallback: OpenRouter

• Environment Variable: OPENROUTER_API_KEY
• Get API Key: Visit OpenRouter
• Note: May experience rate limiting with free tier

API Providers

The system automatically selects the best available AI provider:

Primary: Google Gemini (Recommended)

• Environment Variable: GEMINI_API_KEY
• Get API Key: Visit Google AI Studio
• Advantages: No rate limiting, high quality, cost-effective
• Default Model: gemini-1.5-pro

Fallback: OpenRouter

• Environment Variable: OPENROUTER_API_KEY
• Get API Key: Visit OpenRouter
• Note: May experience rate limiting with free tier

Provider Selection Logic

1. If GEMINI_API_KEY is set → Uses Google Gemini
2. Else if OPENROUTER_API_KEY is set → Uses OpenRouter
3. Otherwise → Throws error requiring API key

Environment Variables

# AI Provider (choose one)
GEMINI_API_KEY=your-gemini-api-key-here
# OR
OPENROUTER_API_KEY=your-openrouter-key-here

# Database
DATABASE_URL="mongodb://localhost:27017/nextlovable"

# Optional overrides
GEMINI_MODEL=gemini-1.5-pro
OPENROUTER_MODEL=openai/gpt-4o

Architecture

• Frontend: Next.js 14 with React and Tailwind CSS
• Backend: Next.js API routes with streaming responses
• Database: MongoDB with Prisma ORM
• Vector Store: ChromaDB with LangChain integration
• AI Framework: LangChain + LangGraph for agent orchestration
• State Management: Custom memory system with conversation persistence

Development

# Install dependencies
pnpm install

# Run in development mode
pnpm run dev

# Build for production
pnpm run build

# Run linting
pnpm run lint

Project Structure

├── app/                    # Next.js app directory
│   ├── api/               # API routes
│   │   ├── chat/         # Main chat endpoint
│   │   └── sessions/     # Session management
│   └── page.tsx          # Main chat interface
├── components/            # React components
│   └── ChatSidebar.tsx   # Project management sidebar
├── lib/                   # Core libraries
│   ├── agents/           # LangGraph agent definitions
│   ├── db/               # Database utilities
│   └── utils/            # Helper functions
├── prisma/                # Database schema
└── src/                   # Source code
    └── agents/           # Agent implementations