Refactory 🔍

⚠️ EXPERIMENTAL ALPHA SOFTWARE

AI-assisted code review tool - Personal experiment in using AI agents for focused code analysis.

Overview

Refactory is a prototype tool that uses AI to review small codebases:

• 🏗️ Architect - Basic design pattern and structure checks
• 🔒 Security - Common vulnerability detection
• ⚡ Performance - Obvious algorithmic issues

Provides scored feedback as a "second pair of eyes" for code review.

Status

⚠️ Personal Research Project: Experimental software for small codebases Works For: Individual files and small projects (~100 files) Experimenting With: Cost reduction through local pattern matching

Quick Start

Installation

git clone <repository-url>
cd refactory
uv venv .venv && uv pip install -e .

Configuration

Create a .env file:

# Recommended: Best value for most users
REFACTORY_MODEL=gemini-2.0-flash-exp
GEMINI_API_KEY=your_api_key_here

Supported Models:

• Google Gemini: gemini-2.0-flash-exp, gemini-1.5-pro, gemini-1.5-flash
• Anthropic Claude: anthropic:claude-4-opus, anthropic:claude-4-sonnet
• OpenAI GPT: openai:gpt-4, openai:gpt-4.1, openai:gpt-4o
• Groq: groq:llama-3.3-70b, groq:deepseek-r1

Basic Usage

# Analyze a single file
refactory analyze examples/sample_code.py

# Analyze a small project
refactory analyze src/ --include "*.py" --exclude ".venv"

# Focus on security issues
refactory analyze auth.py --focus security

# Check multiple areas
refactory analyze code.py --focus security --focus performance

Example Output

🔍 Refactory Analysis Results
┏━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┳━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ File           ┃ Overall Score ┃ Issues ┃ Top Issue                                             ┃
┡━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━╇━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ sample_code.py │      45       │   3    │ Missing input validation                              │
│ auth.py        │      25       │   8    │ SQL injection vulnerability                          │
│ utils.py       │      75       │   1    │ Inefficient algorithm complexity                     │
└────────────────┴───────────────┴────────┴───────────────────────────────────────────────────────┘

What It Does

Basic Capabilities

• Single Files: Analyze individual Python files
• Small Projects: Works best with ~100 files or less
• Focused Analysis: Choose specific areas (security, performance, architecture)
• Simple Output: Table or JSON format

What Each Agent Checks

• 🏗️ Architect: Basic SOLID principles, obvious design issues
• 🔒 Security: Common vulnerabilities like SQL injection, hardcoded secrets
• ⚡ Performance: Obvious inefficiencies, algorithmic complexity issues

Documentation

• CLI Reference - Complete command-line interface guide
• User Guide - Getting started and configuration
• Design Document - Technical architecture and roadmap

Example Use Cases

# Check auth code for security issues
refactory analyze auth.py --focus security

# Review new feature for design issues
refactory analyze new_feature/ --focus architect

# Quick performance check
refactory analyze data_processor.py --focus performance

---

⚠️ Personal research project - use at your own risk.