Skip to content

README.md

Latest commit

 

History

History
170 lines (122 loc) · 4.19 KB

README.md

File metadata and controls

170 lines (122 loc) · 4.19 KB
Raw

🤖 AI Code Reviewer

A full-stack web application that uses Groq AI (Llama 3.3 70b) to automatically review, debug, and fix your code. Simply paste your code, select a programming language, and receive instant feedback including bug detection, explanations, and a fully corrected version of your code.

✨ Features

  • 🧠 AI-Powered Review — Uses Groq's Llama 3.3 70b model to understand and analyze your code like a senior developer
  • 💻 Monaco Editor — The same code editor that powers VS Code, embedded directly in the browser
  • 🐛 Bug Detection — Finds syntax errors, logic errors, and bad practices in your code
  • 🔧 Auto Fix — Provides a fully corrected version of your code with explanations
  • 🌐 Multi-Language Support — Supports Python, JavaScript, Java, C++, and C
  • ⚡ Instant Results — Powered by Groq's ultra-fast inference engine

🛠️ Tech Stack

Layer Technology
Frontend React, Monaco Editor, Axios, React Syntax Highlighter
Backend FastAPI, Python, Uvicorn
AI Model Groq API — Llama 3.3 70b
Styling Inline CSS with VS Code dark theme

📁 Project Structure

ai-code-reviewer/
│
├── backend/
│   ├── main.py            # FastAPI server — handles API routes
│   ├── reviewer.py        # Groq AI logic — sends code to AI and returns review
│   ├── requirement.txt    # Python dependencies
│   └── .env               # API key (not uploaded to GitHub)
│
├── frontend/
│   ├── public/
│   │   └── index.html     # Base HTML file
│   ├── src/
│   │   ├── components/
│   │   │   ├── CodeInput.jsx      # Monaco Editor + language selector + button
│   │   │   └── ReviewOutput.jsx   # Displays AI review with syntax highlighting
│   │   ├── App.jsx        # Main layout — connects all components
│   │   ├── api.js         # Axios API call to backend
│   │   └── index.js       # React entry point
│   └── package.json       # JavaScript dependencies
│
├── .gitignore
└── README.md

🚀 How to Run Locally

Prerequisites

Step 1 — Clone the repository

git clone https://github.com/jonsnow273/code-reviewer.git
cd code-reviewer

Step 2 — Setup Backend

cd backend
pip install -r requirement.txt

Create a .env file inside the backend folder:

GROQ_API_KEY=your-groq-api-key-here

Start the backend server:

uvicorn main:app --reload

Backend will run on http://localhost:8000

Step 3 — Setup Frontend

Open a new terminal:

cd frontend
npm install
npm start

Frontend will run on http://localhost:3000 and open automatically in your browser.

Step 4 — Use the App

  1. Select your programming language from the dropdown
  2. Paste your code into the Monaco Editor
  3. Click "Review Code"
  4. View the AI's analysis on the right panel — bugs found, explanation, and fixed code

🔑 Environment Variables

Variable Description
GROQ_API_KEY Your free Groq API key from console.groq.com

⚠️ Never share your API key or push it to GitHub. Always keep it in the .env file.

📸 How It Works

User pastes code in Monaco Editor
            ↓
Clicks "Review Code" button
            ↓
React sends code + language to FastAPI backend
            ↓
FastAPI calls Groq AI (Llama 3.3 70b)
            ↓
Groq returns bugs found + fixed code + explanation
            ↓
React displays the review with syntax highlighting

🌱 Future Improvements

  • Deploy online (Vercel + Render)
  • Add file upload support (.py, .js, .java)
  • Save review history
  • Add more languages (TypeScript, Rust, Go)
  • Improve UI with Tailwind CSS
  • Add copy button for fixed code

👨‍💻 Author

Pranit — @jonsnow273

📄 License

This project is open source and available under the MIT License.