🤖 AI Agent MVP - Autonomous Software Development

Self-Improving AI Agent for Automatic Software Project Generation

An intelligent AI-powered system that automatically analyzes, designs, codes, tests, and documents complete software projects using state-of-the-art language models via OpenRouter.ai.

🎯 What Does It Do?

The AI Agent MVP takes a simple project description and automatically generates:

📊 Project Analysis - Requirements, user stories, technical specifications
🏗️ Software Architecture - System design, database schema, API structure
💻 Complete Source Code - Ready-to-run project files and folders
🧪 Testing Strategy - Test cases, automation setup, quality gates
📚 Documentation - User guides, API docs, deployment instructions
🔧 Real Project Files - Actual working directories with proper file structure

Input: "Build a todo app with React frontend and Node.js backend"
Output: Complete working project with all files, documentation, and deployment ready!

✨ Key Features

🚀 End-to-End Automation - From idea to working code
🤖 Multi-Model Support - Works with various AI models via OpenRouter
📁 Real File Generation - Creates actual project directories and files
🐳 Docker Ready - Containerized for easy deployment
🔄 Sequential Pipeline - 6-phase structured development process
📊 Comprehensive Logging - Track every step of the generation process
🎯 Production Ready - Clean, documented, and maintainable code output

🛠️ Tech Stack

Component	Technology
Backend	Python 3.12 + FastAPI
AI Integration	OpenRouter.ai API
Current Model	Qwen 3 Coder (Free)
Container	Docker + Docker Compose
Environment	python-dotenv
HTTP Client	requests

🚀 Quick Start

Prerequisites

Docker and Docker Compose
OpenRouter.ai API key (Get yours here)

1. Clone the Repository

git clone https://github.com/yourusername/mvp-ai-agent.git
cd mvp-ai-agent

2. Configure Environment

# Add your OpenRouter API key to .env
echo "OPENROUTER_API_KEY=your_actual_api_key_here" >> .env

3. Build and Run

# Build the Docker image
docker build -t mvp-ai-agent .

# Run with volume mapping for real-time file access
docker run -d -p 8000:8000 -v $(pwd)/outputs:/app/outputs mvp-ai-agent

4. Test the API

# Check if the service is running
curl http://localhost:8000/health

# Generate your first project
curl -X POST "http://localhost:8000/generate" \
  -H "Content-Type: application/json" \
  -d '{"description": "Build a simple todo app with React frontend and Express.js backend"}'

5. Check Generated Files

# View generated project structure
ls -la outputs/
tree outputs/source_code/  # If tree is installed

# Generated files will be in:
# outputs/source_code/frontend/
# outputs/source_code/backend/

📚 API Documentation

Interactive API Docs

Once running, visit: http://localhost:8000/docs

Main Endpoints

`POST /generate`

Generate a complete software project from description.

Request:

{
  "description": "Your project description here"
}

Response:

{
  "success": true,
  "message": "AI agent pipeline completed successfully",
  "execution_time": 45.2,
  "outputs": {
    "analysis": "outputs/analysis.md",
    "design": "outputs/design.md",
    "source_code": "outputs/source_code/implementation.md",
    "testing": "outputs/tests.md",
    "documentation": "outputs/documentation.md"
  },
  "project_generation": {
    "success": true,
    "created_files": ["frontend/App.jsx", "backend/server.js"],
    "created_directories": ["frontend", "backend"],
    "total_files": 8,
    "total_directories": 3
  }
}

Other Endpoints

GET /health - Health check
GET / - Service status

⚙️ Configuration

Environment Variables

Variable	Description	Default
`OPENROUTER_API_KEY`	Your OpenRouter API key	Required
`OPENROUTER_MODEL`	AI model to use	`qwen/qwen3-coder:free`

Model Switching

Available Models:

qwen/qwen3-coder:free (Default - Free)
google/gemini-2.0-flash-exp:free (Free)
anthropic/claude-3.5-sonnet (Paid)
openai/gpt-4o (Paid)

Switch Models:

# Via environment variable
export OPENROUTER_MODEL="google/gemini-2.0-flash-exp:free"

# Via Docker run
docker run -e OPENROUTER_MODEL="anthropic/claude-3.5-sonnet" mvp-ai-agent

📁 Project Structure

mvp-ai-agent/
├── app.py                      # FastAPI orchestrator
├── openrouter_service.py       # AI model integration
├── model_config.py            # Model configuration
├── prompts/                   # AI prompt templates
│   ├── analysis.txt
│   ├── design.txt
│   ├── coding.txt
│   ├── testing.txt
│   └── documentation.txt
├── outputs/                   # Generated content
│   ├── analysis.md           # Project analysis
│   ├── design.md             # Architecture design
│   ├── tests.md              # Testing strategy
│   ├── documentation.md      # Final documentation
│   └── source_code/          # Real project files
│       ├── frontend/         # Frontend application
│       ├── backend/          # Backend application
│       └── database/         # Database files
├── requirements.txt          # Python dependencies
├── Dockerfile               # Container configuration
└── .env                     # Environment variables

🎯 Example Use Cases

Web Applications

{
  "description": "Create a blog platform with user authentication, post creation, comments, and admin panel using Next.js and PostgreSQL"
}

Mobile Backend

{
  "description": "Build a REST API for a fitness tracking app with user profiles, workout logging, and progress analytics using FastAPI and MongoDB"
}

Desktop Application

{
  "description": "Develop a file organizer desktop app with drag-drop interface, automatic categorization, and cloud sync using Electron and Node.js"
}

🔄 Development Workflow

The 6-Phase Pipeline

🔍 Analysis Phase - Requirements gathering, user stories, technical specs
🏗️ Design Phase - System architecture, database design, API specs
💻 Coding Phase - Complete source code, project structure, config files
🧪 Testing Phase - Test strategies, test cases, quality assurance
📚 Documentation Phase - User docs, API docs, deployment guides
🔧 Real Project Generation - Actual file creation, directory structure

📊 Performance & Limitations

Metric	Value
Average Generation Time	Depending on the performance of the language model
Supported File Types	JavaScript, Python, HTML, CSS, JSON, Markdown
Max Project Complexity	Medium to large applications
Concurrent Requests	Limited by OpenRouter rate limits

🤝 Contributing

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

Development Setup

# Clone and setup
git clone https://github.com/yourusername/mvp-ai-agent.git
cd mvp-ai-agent

# Install dependencies
pip install -r requirements.txt

# Run locally
python app.py

📄 License

This project is licensed under the MIT License.

🙏 Acknowledgements

OpenRouter.ai for AI model access
FastAPI for the excellent web framework
Qwen Team for the coding-specialized model

📞 Support

🐛 Issues: GitHub Issues
💬 Discussions: GitHub Discussions

Made with ❤️ by Ömer Yasir Önal(https://github.com/yourusername)

"From idea to implementation in minutes, not months."

Name		Name	Last commit message	Last commit date
Latest commit History 8 Commits
prompts		prompts
.DS_Store		.DS_Store
.gitignore		.gitignore
Dockerfile		Dockerfile
README.md		README.md
app.py		app.py
model_config.py		model_config.py
openai_service.py		openai_service.py
openrouter_service.py		openrouter_service.py
requirements.txt		requirements.txt

OmerYasirOnal/Agent-MVP

Folders and files

Latest commit

History

Repository files navigation