Code Reviewer with AutoGen

This project demonstrates how to build a simple multi-agent system using AutoGen where:

• A Tutor Agent writes Python code based on user queries.
• A Reviewer Agent critiques the code for correctness, readability, and improvements.
• A Student Proxy Agent manages the conversation flow, executes code safely in a sandbox, and stores logs.

The system also:

• Saves the entire chat transcript into chat_log.txt.
• Extracts the Tutor’s final code block and saves it into final_code.txt for testing or reuse.

---

🚀 Features

• Tutor Agent: Generates clear, correct Python solutions.
• Reviewer Agent: Reviews Tutor’s solutions and suggests improvements.
• Student Proxy Agent: Automates interaction, prevents infinite loops, and runs code in a sandbox.
• Conversation Logging: Full transcript saved to chat_log.txt.
• Code Extraction: Clean Python code saved to final_code.txt.

---

📂 Project Structure

code-reviewer-autogen/
│── sandbox/             # Safe environment for code execution
│── chat\_log.txt         # Full transcript of Tutor & Reviewer conversation
│── final\_code.txt       # Extracted Python code from Tutor
│── main.py              # Main script (agents + logic)
│── requirements.txt     # Python dependencies
│── .gitignore
│── README.md

---

⚙️ Installation

1. Clone the repository:

   git clone https://github.com/YOUR-USERNAME/code-reviewer-autogen.git
   cd code-reviewer-autogen

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate    # macOS/Linux
   venv\Scripts\activate       # Windows

3. Install dependencies:

   pip install -r requirements.txt

---

🔑 OpenAI API Key

This project requires an OpenAI API key. The script loads it from system environment variables using:

import os

config = {
    "model": "gpt-4o-mini",
    "api_key": os.getenv("OPENAI_API_KEY")
}

Setting the key in system variables:

• macOS/Linux (bash/zsh):

  export OPENAI_API_KEY="your_api_key_here"

• Windows (PowerShell):

  setx OPENAI_API_KEY "your_api_key_here"

Alternative: Set directly in code (not recommended for production)

If you don’t want to use environment variables, you can hardcode it:

config = {
    "model": "gpt-4o-mini",
    "api_key": "your_api_key_here"
}

---

▶️ Usage

Run the main script:

python main.py

After execution:

• chat_log.txt will contain the full conversation (Tutor + Reviewer).
• final_code.txt will contain the clean extracted Python code.

---

🛠️ Example Flow

1. Student asks Tutor: "Write a function that prints n factorial for n to be any number"
2. Tutor generates the Python code.
3. Student executes it inside the sandbox.
4. Reviewer critiques the Tutor’s code and suggests improvements.
5. Logs and final code are saved for review.

---

📌 Future Improvements

• Add a Critic Agent to review the Reviewer’s feedback.
• Integrate with linters (e.g., Pylint/Flake8) as tools for Reviewer.
• Extend pipeline to support multiple rounds of refinement.
• Store logs in a database instead of flat files.