Live Demo: https://code-critic-six.vercel.app
CodeCritic is a full-stack AI-powered code review platform designed to help developers write cleaner, faster, and more reliable code. The platform leverages Groq's LLaMA 3.3 70B model to perform intelligent code analysis, detect bugs, identify performance bottlenecks, suggest optimizations, and calculate time and space complexity across multiple programming languages.
Users can analyze code instantly, receive structured AI feedback, optimize code with a single click, manage review history, and improve code quality through actionable recommendations powered by advanced AI.
- Analyze code using LLaMA 3.3 70B via Groq API.
- Detect syntax errors, logic bugs, runtime issues, and edge cases.
- Receive structured feedback with detailed explanations.
- Identify security vulnerabilities and coding best-practice violations.
- One-click code optimization.
- Automatically fixes bugs and edge cases.
- Improves performance and readability.
- Generates production-ready optimized code.
- Compare complexity before and after optimization.
- Monaco Editor integration for a VS Code-like experience.
- Syntax highlighting for multiple programming languages.
- Real-time language detection.
- Clean and distraction-free development workflow.
- Automatically save every code review session.
- Browse previous code analyses.
- Reload and inspect past reviews.
- Delete unwanted sessions anytime.
- Supports all major programming languages.
- Automatic language detection powered by AI.
- Language-specific analysis and optimization.
- Consistent review experience across technologies.
- Clerk-based authentication and user management.
- Email/password authentication.
- Google OAuth integration.
- GitHub OAuth integration.
- Manage user profile information.
- Upload custom profile avatars.
- Update account details securely.
- Password management and account controls.
- Built using React and TailwindCSS.
- Premium dark-themed developer experience.
- Fully responsive across desktop and mobile devices.
- Clean and intuitive interface inspired by modern SaaS platforms.
- React 19
- Vite
- TailwindCSS
- React Router
- Monaco Editor
- Axios
- Clerk React
- React Icons
- DevIcons React
- Node.js
- Express.js
- MongoDB Atlas
- Mongoose ODM
- Clerk Authentication
- Groq SDK
- Svix Webhooks
- Helmet
- CORS
- Morgan
- Groq API
- LLaMA 3.3 70B Versatile
- AI Code Review
- AI Code Optimization
- Programming Language Detection
- Complexity Analysis
- Frontend: Vercel
- Backend: Render
- Database: MongoDB Atlas
Before running the project, ensure you have:
- Node.js installed
- MongoDB Atlas account
- Clerk account
- Groq API key
git clone https://github.com/karunya666/CodeCritic.gitcd backend
npm installcd frontend
npm installcd backend
npm run devcd frontend
npm run devMONGODB_URI=your_mongodb_connection_string
GROQ_API_KEY=your_groq_api_key
CLERK_SECRET_KEY=your_clerk_secret_key
CLERK_WEBHOOK_SECRET=your_clerk_webhook_secret
CLIENT_URL=http://localhost:5173
PORT=5000
NODE_ENV=developmentVITE_API_URL=http://localhost:5000/api
VITE_CLERK_PUBLISHABLE_KEY=your_clerk_publishable_key| Route | Description |
|---|---|
| POST /api/review | Analyze code and create review session |
| POST /api/review/optimise | Generate optimized code |
| POST /api/review/detect | Detect programming language |
| PATCH /api/review/:id | Update an existing review session |
| GET /api/sessions | Retrieve user review history |
| GET /api/sessions/:id | Get review session details |
| DELETE /api/sessions/:id | Delete review session |
| POST /api/webhooks/clerk | Sync Clerk user data |
- User writes or pastes code into the Monaco Editor.
- Code is submitted for AI analysis.
- Groq's LLaMA 3.3 70B model reviews the code.
- AI identifies errors, improvements, and complexity metrics.
- Structured feedback is displayed in the review panel.
- Review session is automatically stored for future access.
- User clicks the Optimize button.
- AI rewrites the code using best practices.
- Bugs and inefficiencies are removed.
- Optimized code replaces the original version.
- Code is automatically re-analyzed.
- Updated complexity metrics and recommendations are displayed.
https://code-critic-six.vercel.app
https://codecritic-ydsh.onrender.com
MongoDB Atlas
- Algorithm Visualizer with step-by-step execution.
- Export code reviews as PDF.
Karunya
Built to help developers write better code, identify bugs faster, optimize performance, and improve software quality through AI-powered code reviews.