policyguard-paolovella 0.1.0

Pre-commit hook for detecting security issues in AI-generated code

🛡️ PolicyGuard

Pre-commit security scanner for AI-generated Python code

---

PolicyGuard catches common security vulnerabilities in your Python code, especially those often introduced by AI coding assistants like GitHub Copilot.

✨ Features

• 🚨 Detects hardcoded secrets - passwords, API keys, tokens
• 💉 Prevents SQL injection - finds string concatenation in queries
• 🎯 Catches dangerous functions - eval(), exec(), pickle.load()
• 🤖 AI-specific patterns - TODOs for passwords, example keys
• ⚡ Fast - typically <100ms per file
• 🎨 Beautiful output - color-coded, with context
• 🔧 Git pre-commit hook - catch issues before they're committed
• 📦 Zero dependencies - minimal requirements

🚀 Quick Start

Install

pip install policyguard

Initialize in your repo

cd your-python-project
policyguard init

That's it!

PolicyGuard will now check your Python files before each commit.

📋 What It Catches

Security Issues

• Hardcoded passwords and secrets
• SQL injection vulnerabilities
• Command injection risks
• Insecure deserialization (pickle)
• Weak random number generation
• Dangerous eval/exec usage

AI-Specific Patterns

• # TODO: change password comments
• Example API keys (sk-xxx, AKIA-xxx)
• Placeholder emails ([e-mail])
• Localhost URLs in production code

🔧 Usage

Scan manually

# Scan current directory
policyguard scan

# Scan specific file
policyguard scan auth.py

# Output as JSON
policyguard scan --json

View all rules

policyguard rules

🤝 Contributing

We love contributions! Please see our contributing guidelines.

📄 License

MIT - see LICENSE

---

Built with ❤️ to keep AI-generated code secure