Full-stack code generation platform. Describe what you want, get production-ready code with tests and docs.
Generates code in 10 languages (Python, JavaScript, TypeScript, Java, C#, Go, Rust, C++, Ruby, Swift) using OpenAI's GPT-4. Includes automatic test generation with framework detection and inline documentation. Users provide their own API key stored locally in browser.
cd backend
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate
pip install -r requirements.txt
uvicorn main:app --reload --port 8000cd frontend
npm install
npm run dev- Add your OpenAI API key (stored locally in browser)
- Enter what you want to build
- Select programming language
- Generate code, tests, and documentation
- Copy or download results
Frontend (Next.js + TypeScript)
↓ HTTP + Auth Header
Backend (FastAPI)
↓ User's API Key
OpenAI GPT-4
User's OpenAI API key stored in localStorage, sent in Authorization header. Backend creates separate LangChain instance per request with provided key.
POST /api/generate- Generate code (requiresAuthorization: Bearer <api_key>)POST /api/analyze- Analyse code qualityGET /api/languages- Supported languagesGET /api/health- Health check
Backend (.env):
# Optional - users provide their own
OPENAI_API_KEY=""
API_HOST=0.0.0.0
API_PORT=8000
CORS_ORIGINS=["http://localhost:3000"]Frontend (.env.local):
NEXT_PUBLIC_API_URL=http://localhost:8000/apiBackend: FastAPI, LangChain, OpenAI API, tree-sitter, Pydantic Frontend: Next.js 15, React 19, TypeScript, Tailwind CSS, Framer Motion
- Multi-language support with framework-specific test generation
- Code quality metrics (complexity, readability, performance)
- Syntax validation using tree-sitter AST parsing
- Matrix-themed cyberpunk UI
- No API key storage on backend - users bring their own
Tests: cd backend && pytest
Linting: cd frontend && npm run lint
See CLAUDE.md for detailed development guide.
MIT