CodeCraft AI

AI-powered code review assistant that provides instant, detailed feedback on code quality, security, and best practices

Comment

Inspiration

Every developer has experienced the frustration of writing code that works but could be better. Traditional code reviews are time-consuming, inconsistent, and limited by access to senior developers. As developers ourselves, we wanted to democratize quality code feedback - making it available to everyone, instantly, 24/7.

We were inspired to create CodeCraft AI to solve this problem. By leveraging the power of modern AI, we built a platform that acts as a personal Senior Software Engineer, providing instant, educational code reviews across multiple programming languages.

What it does

CodeCraft AI is an AI-powered code review assistant that helps developers of all levels learn, improve, and write better code. Users paste their code into our editor, select their programming language, and receive instant, detailed feedback in four key areas:

  • 🔍 Logic Analysis: Detects bugs, edge cases, and logical flaws
  • ⚡ Efficiency Review: Identifies performance bottlenecks and optimization opportunities
  • 🎨 Code Style Check: Ensures clean, readable, and maintainable code
  • 🔒 Security Audit: Finds vulnerabilities and security risks

The feedback is educational - explaining why something is wrong and how to fix it. Users can save their review history to track their progress over time.

How we built it

Backend:

  • FastAPI (Python): Built a high-performance REST API with automatic OpenAPI documentation
  • PostgreSQL + SQLAlchemy: Designed a robust database schema for users and review history
  • JWT Authentication: Implemented secure user authentication with bcrypt password hashing
  • OpenAI GPT-4: Engineered sophisticated prompts to get structured, educational feedback
  • Docker: Containerized the entire application for consistent development and deployment

Frontend:

  • Next.js 14 + React 18: Built a responsive, modern single-page application
  • Tailwind CSS: Created a beautiful, custom UI with utility-first styling
  • react-simple-code-editor + Prism.js: Implemented syntax-highlighted code editing across 10+ languages
  • Axios + React Context: Managed API calls and global authentication state

Infrastructure:

  • Docker Compose: Orchestrated multi-container setup (PostgreSQL, Backend, Frontend)
  • GitHub Actions: Set up CI/CD pipelines for automated testing

🔗 Deployment & Documentation

Live Demo:
👉 https://codecraft-ai-one.vercel.app/ (Deployed on Vercel.com - try it now!)

API Documentation:
📖 https://github.com/abdulboyprogramming-arch/codecraft-ai/blob/main/docs/API.md (Complete API reference with all endpoints)

Source Code:
💻 https://github.com/abdulboyprogramming-arch/codecraft-ai (Full source code available on GitHub)

The application is fully containerized and deployed on Render.com, featuring:

  • ✅ Automatic deployments from GitHub
  • ✅ PostgreSQL database on Render
  • ✅ Scalable web services
  • ✅ Built-in health checks and monitoring

Challenges we ran into

1. Prompt Engineering for Quality Feedback: Getting GPT-4 to provide consistently structured, educational feedback was a challenge. We iterated extensively on our system prompts to ensure the AI would respond with categorized, actionable feedback in JSON format.

2. Dependency Management: Managing compatibility between Python packages (FastAPI, SQLAlchemy, OpenAI, LangChain) across different versions required careful version pinning and conflict resolution.

3. Real-time User Experience: Designing a seamless experience where users can paste code and instantly receive feedback required careful state management and loading states in the React frontend.

4. Windows Development Environment: Overcame challenges with Python 3.14 compatibility (using Python 3.11 instead) and managed virtual environment setups across development machines.

5. Favicon/Logo Generation: Created a professional, scalable favicon that works at all sizes from 16px to 512px while representing the brand identity.

6. Deployment Optimization: Optimized the application for deployment on Render.com, managing environment variables and configuring the production build process.

Accomplishments that we're proud of

  • ✅ Built a complete, production-ready full-stack application in record time
  • ✅ Integrated OpenAI's GPT-4 to provide meaningful, educational code feedback
  • ✅ Created a beautiful, responsive UI that works on desktop and mobile
  • ✅ Implemented secure authentication with JWT and password hashing
  • ✅ Containerized the entire application with Docker for easy deployment
  • ✅ Wrote comprehensive documentation for API, architecture, and deployment
  • ✅ Achieved ~80% test coverage with both backend and frontend tests
  • ✅ Successfully deployed the application to Render.com with CI/CD

What we learned

  • Prompt Engineering: How to craft effective prompts to get structured, educational AI responses
  • Full-Stack Development: Integrating FastAPI with Next.js, managing state, and handling API calls
  • Dependency Management: The importance of version compatibility in Python and Node.js ecosystems
  • DevOps: Setting up Docker, CI/CD pipelines, and managing environment variables
  • User Experience: Designing a clean, intuitive interface for developers
  • Security: Implementing JWT authentication, password hashing, and secure API design
  • Cloud Deployment: Deploying full-stack applications to Render.com with managed databases

What's next for CodeCraft AI

Short-term:

  • 🚀 VS Code Extension for in-IDE code reviews
  • 🤖 GitHub Integration for automated PR reviews
  • 📱 Team collaboration features and shared workspaces
  • 🎨 Advanced AI feedback with code correction suggestions

Medium-term:

  • 🧠 Fine-tuned AI models specialized for specific languages
  • 📊 Analytics dashboard for code quality metrics
  • 🔗 CI/CD pipeline integration (GitHub Actions, Jenkins)
  • 📱 Mobile app for on-the-go code review

Long-term:

  • 🌐 Global scaling with multi-region deployment
  • 🏢 Enterprise features: SSO, audit logging, SOC 2 compliance
  • 🔮 Predictive code analysis and bug prevention
  • 🎯 AI-powered code completion and test generation

📊 Project Links

Resource Link
Live Demo https://codecraft-ai-one.vercel.app/
GitHub Repository https://github.com/abdulboyprogramming-arch/codecraft-ai
API Documentation https://github.com/abdulboyprogramming-arch/codecraft-ai/blob/main/docs/API.md
Architecture Docs https://github.com/abdulboyprogramming-arch/codecraft-ai/blob/main/docs/ARCHITECTURE.md
Deployment Guide https://github.com/abdulboyprogramming-arch/codecraft-ai/blob/main/docs/DEPLOYMENT.md

Try CodeCraft AI today and experience the future of code reviews! 🚀

Built With

Share this project:

Updates

posted an update —

Hey everyone!

I'm Abdulrahman, a software engineering student at Federal University of Technology, Akure (FUTA). CodeCraft AI is my submission for the Prometheus July AI Challenge 2026.

The inspiration came from my own experience as a developer. I've often wished I could get instant, reliable feedback on my code without having to wait for a busy senior developer. So I built CodeCraft AI—a tool that acts like a personal Senior Software Engineer, available 24/7.

It uses OpenAI's GPT-4 to analyze code and provide structured, educational feedback. It's not just about finding bugs—it's about learning why something is wrong and how to fix it.

  • 5+ languages supported
  • Logic, Efficiency, Style & Security analysis
  • Full review history
  • Completely free to use

Check it out and let me know what you think!

Live Demo: https://codecraft-ai-one.vercel.app GitHub: https://github.com/abdulboyprogramming-arch/codecraft-ai

Built with: FastAPI, Next.js, PostgreSQL, OpenAI API, Docker, Tailwind CSS

Log in or sign up for Devpost to join the conversation.

Submission history