🚀 Multimodal Coding Agent with Gemini

A powerful multimodal coding assistant that can analyze images containing code problems and generate solutions in multiple programming languages. Built with Google's Gemini AI and Gradio for an intuitive web interface.

✨ Features

🔍 Multimodal Analysis

• Image Processing: Upload screenshots of coding problems or code snippets
• Text Input: Describe coding problems in natural language
• Smart Recognition: Automatically detects problem types and requirements

💻 Multi-Language Support

• Python: Full execution with safety restrictions
• JavaScript/TypeScript: Modern ES6+ and TypeScript support
• Java: Complete compilation and execution
• C/C++: GCC/G++ compiler integration
• HTML/CSS: Syntax validation and web project generation
• TSX: React TypeScript components

🌐 Web Project Generation

• Complete Websites: Generate full HTML, CSS, and JavaScript projects
• Responsive Design: Mobile-friendly, modern web standards
• Auto-Save: Automatically organize and save multi-file projects
• Ready-to-Deploy: Generated projects work out of the box

🛡️ Safe Code Execution

• Sandboxed Environment: Restricted execution for security
• Module Filtering: Only allows safe standard library modules
• Timeout Protection: Prevents infinite loops and hanging processes
• Error Handling: Comprehensive error reporting and suggestions

🚀 Quick Start

Prerequisites

• Python 3.8 or higher
• Google Gemini API key
• (Optional) Compilers for specific languages:
  • Node.js for JavaScript/TypeScript
  • JDK for Java
  • GCC/G++ for C/C++

Installation

1. Clone the repository

   git clone <your-repo-url>
   cd yehh

2. Install dependencies

   pip install -r requirements.txt

3. Get your Gemini API key

   • Visit Google AI Studio
   • Create a new API key
   • Keep it ready for the application

4. Run the application

   python app.py

5. Open your browser

   • Navigate to the URL shown in the terminal (typically http://localhost:7860)
   • Enter your Gemini API key
   • Start coding!

💡 Usage Examples

📸 Image Analysis

1. Upload an image containing a coding problem
2. Select your target programming language
3. Click "Generate & Execute Solution"
4. Get code generation + execution results

📝 Text Prompts

- "Write a function to reverse a string in Python"
- "Create a responsive login form using HTML, CSS, and JS"
- "Build a calculator class in Java"
- "Make a MVP landing page"

🌐 Web Projects

Input: "Create a modern portfolio website"
Output: Complete project with:
├── index.html (Semantic HTML5)
├── styles.css (Responsive CSS)
└── script.js (Interactive JavaScript)

🔧 Advanced Features

Multi-File Project Keywords

Trigger automatic web project generation with keywords:

• landing page, website, web app
• MVP, portfolio, dashboard
• html css js, full website

Language-Specific Execution

• Python: Executes safely with restricted imports
• JavaScript: Requires Node.js installation
• Java: Compiles and runs with JDK
• C/C++: Compiles with GCC/G++ and executes
• HTML/CSS: Validates syntax and structure

Error Handling

• API Errors: Graceful handling of service unavailability
• Compilation Errors: Clear error messages and fix suggestions
• Runtime Errors: Detailed execution analysis and debugging help

🏗️ Project Structure

yehh/
├── app.py                 # Main application file
├── requirements.txt       # Python dependencies
├── README.md             # Project documentation
├── LICENSE               # MIT License
├── .gitignore           # Git ignore patterns
├── image.png            # Sample image for testing
└── generated_project/   # Auto-generated web projects
    ├── index.html
    ├── styles.css
    └── script.js

🔒 Security Features

• Restricted Execution: Limited to safe operations only
• Module Filtering: Prevents dangerous imports
• Timeout Controls: Automatic termination of long-running code
• File System Protection: No unauthorized file access

🤝 Contributing

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

🐛 Troubleshooting

Common Issues

503 Service Unavailable Error

• Gemini API is temporarily down
• Wait a few minutes and retry
• Use text input instead of image upload

Import/Module Errors

• Run pip install -r requirements.txt to install all dependencies
• If you get google-genai errors, run: pip install google-genai
• For agno installation issues, try: pip install agno --upgrade

Execution Errors

• Ensure required compilers are installed
• Check language-specific prerequisites
• Review generated code for syntax errors

API Key Issues

• Verify your Gemini API key is valid
• Check API quotas and limits
• Ensure vision API access is enabled

📋 Roadmap

• Support for more programming languages (Go, Rust, etc.)
• Advanced code optimization suggestions
• Integration with popular IDEs
• Collaborative coding features
• Custom execution environments
• API endpoint for programmatic access

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Google Gemini: For powerful multimodal AI capabilities
• Gradio: For the intuitive web interface
• Agno: For seamless AI agent integration
• Community: For testing and feedback

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: [email protected]

---

**Made with ❤️ by hari7261

Transform your coding workflow with AI-powered multimodal assistance!