# 🤖 AI Code Reviewer

A **production-level AI-powered code reviewer** that analyzes code across multiple languages, provides intelligent feedback, generates improved code suggestions, and tracks user history — all through an interactive web interface.

---

## 🚀 Features

### 🧠 AI-Powered Code Review

* Uses local LLM (Ollama) for intelligent analysis
* Detects bugs, code smells, and best practice violations
* Provides structured feedback and suggestions

### ⚙️ Hybrid Scoring System

* Combines:

  * AI-based evaluation (60%)
  * Rule-based static analysis (40%)
* Produces a final **code quality score**

### 🌐 Multi-Language Support

* Python
* JavaScript
* Java
* C++

### 🛠️ Auto Code Improvement

* Generates improved versions of code
* Maintains original logic
* Provides multiple variations (regenerate option)

### 👤 User Authentication

* Simple login system
* Session-based access control

### 📜 History Tracking

* Stores past analyses per user
* Displays recent activity
* Downloadable history reports

### 📥 Export Features

* Download improved code
* Download analysis history

### 🖥️ Interactive UI

* Built with Streamlit
* Clean, responsive interface
* Real-time feedback

---

## 🏗️ Project Structure

```bash
ai-code-reviewer/
│
├── app.py
│
├── analyzer/
│   ├── ai_review.py
│   └── static_analysis.py
│
├── parser/
│   └── code_parser.py
│
├── utils/
│   ├── formatter.py
│   └── auth_storage.py
│
├── data/
│   └── history/
│
├── requirements.txt
└── README.md
```

---

## ⚙️ Installation

### 1️⃣ Clone Repository

```bash
git clone https://github.com/YOUR_USERNAME/ai-code-reviewer.git
cd ai-code-reviewer
```

---

### 2️⃣ Install Dependencies

```bash
pip install -r requirements.txt
```

---

### 3️⃣ Setup Environment Variables

```bash
export JARVIS_USER=admin
export JARVIS_PASS=1234
```

(For Windows PowerShell:)

```powershell
setx JARVIS_USER "admin"
setx JARVIS_PASS "1234"
```

---

### 4️⃣ Start Ollama (AI Engine)

```bash
ollama run phi3
```

---

### 5️⃣ Run Application

```bash
streamlit run app.py
```

---

## 🧪 Usage

1. Login using credentials
2. Upload a code file
3. Click **Analyze Code**
4. View:

   * Static analysis
   * AI feedback
   * Hybrid score
5. Generate improved code
6. Download results or view history

---

## 📊 Example Output

```text
AI Score: 7/10
Rule Score: 6/10
Final Score: 6.6/10

Issues:
- Missing error handling
- Poor variable naming

Suggestions:
- Improve naming conventions
- Add try-except blocks
```

---

## 🧠 Tech Stack

* **Frontend:** Streamlit
* **Backend:** Python
* **AI Model:** Ollama (Local LLM)
* **Static Analysis:** AST (Python)
* **Automation:** Custom rule engine

---

## 🔐 Authentication

* Lightweight session-based login
* Credentials via environment variables
* Per-user data isolation

---

## 📌 Key Highlights

* Combines **AI + rule-based analysis**
* Supports **multiple programming languages**
* Includes **auto code generation & improvement**
* Tracks **user activity history**
* Designed with **modular architecture**

---

## 🚀 Future Improvements

* Database integration (MongoDB / SQLite)
* Advanced authentication (JWT)
* CI/CD integration
* Code diff visualization
* Team collaboration features

---

## 👨‍💻 Author

**Anirodh Padhy**

GitHub: https://github.com/Aniordh-Padhy
LinkedIn: www.linkedin.com/in/anirodh-padhy-ab3455315

---

## ⭐ Support

If you like this project:

* ⭐ Star the repo
* 🍴 Fork it
* 📢 Share it

---

## 🏁 Conclusion

This project demonstrates a **real-world AI developer tool** combining:

* Machine Learning
* Software Engineering
* Automation
* User Experience

Built with a focus on **performance, usability, and scalability** 🚀