A powerful web application that uses multiple AI models to generate and review code. This application leverages ChatGPT-5 for code generation, and uses both Claude Sonnet 4.5 and Gemini 2.5 for comprehensive code reviews.
- AI-Powered Code Generation: Describe your application requirements and let ChatGPT generate the complete code
- Multi-Model Code Review: Get code reviews from both Claude Sonnet 4.5 and Gemini 2.5
- Intelligent Evaluation: ChatGPT evaluates the reviews and provides recommendations for improvements
- Modern UI: Clean, responsive interface with real-time feedback
- Comprehensive Workflow: Seamlessly orchestrates multiple AI models for optimal results
- React 18: Modern React application with hooks
- Responsive Design: Works on desktop and mobile devices
- Real-time Updates: Live feedback during code generation and review
- Node.js/Express: RESTful API server
- Multi-AI Integration:
- OpenAI GPT-4 Turbo (preparing for GPT-5)
- Anthropic Claude Sonnet 4.5
- Google Gemini 2.5
- Parallel Processing: Reviews run simultaneously for faster results
- Node.js 16.x or higher
- npm or yarn
- API keys for:
- OpenAI (ChatGPT)
- Anthropic (Claude)
- Google (Gemini)
git clone <repository-url>
cd ServerTracking# Install backend dependencies
npm install
# Install frontend dependencies
cd client
npm install
cd ..Create a .env file in the root directory:
cp .env.example .envEdit .env and add your API keys:
# API Keys for AI Models
OPENAI_API_KEY=your_openai_api_key_here
ANTHROPIC_API_KEY=your_anthropic_api_key_here
GOOGLE_API_KEY=your_google_api_key_here
# Server Configuration
PORT=3001
NODE_ENV=development- Visit https://platform.openai.com/api-keys
- Sign in or create an account
- Click "Create new secret key"
- Copy the key and add to
.env
- Visit https://console.anthropic.com/
- Sign in or create an account
- Navigate to API Keys section
- Create a new key and add to
.env
- Visit https://makersuite.google.com/app/apikey
- Sign in with your Google account
- Create an API key
- Copy and add to
.env
Run both frontend and backend concurrently:
npm run devOr run them separately:
# Terminal 1 - Backend
npm run server
# Terminal 2 - Frontend
npm run client# Build the frontend
npm run build
# Start the server
npm startThe application will be available at:
- Frontend: http://localhost:3000
- Backend API: http://localhost:3001
- Enter your application requirements in the text area
- Click "Generate Application"
- Wait for ChatGPT to generate the code
- View the generated code in the output section
- After generating code, click the "Code Review" button in the header
- The system will:
- Send the code to Claude Sonnet 4.5 for review
- Send the code to Gemini 2.5 for review (runs in parallel)
- Send both reviews to ChatGPT for evaluation
- View individual reviews from Claude and Gemini
- Read ChatGPT's evaluation and recommendations
Generate application code from requirements.
Request:
{
"requirements": "Create a todo list application with..."
}Response:
{
"success": true,
"generatedCode": "...",
"model": "gpt-4-turbo-preview",
"usage": {...}
}Review code using Claude and Gemini, then evaluate with ChatGPT.
Request:
{
"code": "function example() {...}"
}Response:
{
"success": true,
"reviews": {
"claude": {...},
"gemini": {...}
},
"evaluation": "..."
}Complete workflow: generate, review, and evaluate in one request.
Request:
{
"requirements": "Create a todo list application..."
}ServerTracking/
├── client/ # React frontend
│ ├── public/
│ ├── src/
│ │ ├── components/
│ │ ├── services/
│ │ │ └── api.js # API client
│ │ ├── App.js # Main component
│ │ ├── App.css # Styles
│ │ └── index.js # Entry point
│ └── package.json
├── server/ # Express backend
│ ├── controllers/
│ │ └── appController.js
│ ├── routes/
│ │ └── appRoutes.js
│ ├── services/
│ │ ├── openaiService.js
│ │ ├── claudeService.js
│ │ └── geminiService.js
│ └── index.js # Server entry point
├── .env.example # Environment template
├── .gitignore
├── package.json
└── README.md
- User Input: User describes desired application features
- Code Generation: ChatGPT generates complete application code
- Code Review:
- Claude Sonnet 4.5 analyzes code for quality, bugs, and best practices
- Gemini 2.5 provides additional review (runs in parallel)
- Evaluation: ChatGPT evaluates both reviews and determines:
- Which suggestions are valid
- Which improvements should be incorporated
- Provides refined recommendations
- Diverse Perspectives: Different AI models catch different issues
- Comprehensive Review: Multiple viewpoints ensure thorough analysis
- Quality Assurance: ChatGPT validates and consolidates feedback
- Best Practices: Combined expertise from three leading AI models
- Ensure all API keys are correctly set in
.env - Check that keys have sufficient credits/quota
- Verify keys are not expired
- If port 3000 or 3001 is in use, update in
.envandclient/package.json
- Ensure backend is running on the correct port
- Check proxy configuration in
client/package.json
- Never commit
.envfile to version control - Keep API keys secure and private
- Use environment variables for all sensitive data
- Consider implementing rate limiting for production use
- Save generated applications to files
- Export code in different formats
- User authentication and project management
- Version history for generated code
- Direct code editing with real-time review
- Support for additional AI models
- Automated testing generation
Contributions are welcome! Please follow these steps:
- Fork the repository
- Create a feature branch
- Make your changes
- Submit a pull request
MIT License - see LICENSE file for details
For issues or questions:
- Open an issue on GitHub
- Check existing documentation
- Review API provider documentation
Built with ChatGPT, Claude Sonnet 4.5, and Gemini 2.5