ai-agent-debuggy-autodev

# AI Code Assistant

A CLI tool that uses Gemini AI to debug and fix Python code. It scans files, modifies code, and tests fixes autonomously.

## Features

- Accepts coding tasks via CLI
- Uses Gemini API to reason and act
- Can:
  - Scan project files
  - Read/write code
  - Run Python scripts
  - Execute tests and verify changes
- Repeats actions until task is done or fails
- Security-focused with sandboxed execution
- Optimized performance with efficient file operations

## Requirements

- Python 3.10+
- A Claude API key from [Google AI Studio](https://aistudio.google.com/)

## Installation

### 1. Clone the Repository

```bash
git clone https://github.com/predrag-milanovic/ai-agent-debuggy-autodev.git
cd ai-agent-debuggy-autodev
```

### 2. Set Up Virtual Environment

```bash
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
```

### 3. Install Dependencies

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

### 4. Configure API Key

Create a `.env` file in the project root:

```bash
echo 'GEMINI_API_KEY="your_key_here"' > .env
```

Replace `your_key_here` with your actual Claude API key from [Google AI Studio](https://aistudio.google.com/).

## Configuration

The project can be configured via environment variables or by editing `config.py`:

- **`AI_WORKING_DIR`** - The working directory for code operations (default: `./calculator`)
- **`AI_MAX_ITERS`** - Maximum number of AI iterations (default: `20`)
- **`MAX_CHARS`** - Maximum characters to read from files (default: `10000`)

Example using environment variables:

```bash
export AI_WORKING_DIR="./my-project"
export AI_MAX_ITERS="30"
```

## Usage

### Basic Usage

```bash
python main.py "your prompt here"
```

### Examples

**Debug and fix code:**
```bash
python main.py "Fix the calculator to handle division by zero"
```

**Add new features:**
```bash
python main.py "Add support for square root operations"
```

**Run with verbose output:**
```bash
python main.py "Check if all tes