zscole/ai-poc-daily/python-security-review

Python Security Code Review

Overview

Review Python code for security vulnerabilities, focusing on the OWASP Top 10 and Python-specific issues.

Critical Checks (Always Flag)

1. Injection Vulnerabilities

• SQL: Look for string formatting in queries

BAD: f"SELECT * FROM users WHERE id = {user_id}" GOOD: cursor.execute("SELECT * FROM users WHERE id = ?", (user_id,))

• Command Injection: subprocess with shell=True

BAD: subprocess.run(f"ls {user_input}", shell=True) GOOD: subprocess.run(["ls", user_input])

• SSTI: Direct user input in templates

BAD: render_template_string(user_input)

2. Authentication Issues

• Hardcoded credentials
• Weak password hashing (MD5, SHA1 without salt)
• Missing rate limiting on auth endpoints
• JWT without expiration

3. Data Exposure

• Secrets in code or logs
• Debug mode in production
• Verbose error messages to users
• Sensitive data in URLs

4. Insecure Deserialization

• pickle.loads with untrusted data
• yaml.load without Loader=SafeLoader
• eval/exec on user input

5. Path Traversal

• User input in file paths without validation
• Missing path canonicalization

Output Format

For each issue found:

[SEVERITY] Issue Title
Line: X
Code: `affected code snippet`
Risk: What could happen
Fix: How to fix it

Severity levels: CRITICAL, HIGH, MEDIUM, LOW, INFO

False Positive Prevention

• Check if input is sanitized before flagging
• Look for middleware/decorators that might handle security
• Consider the context (internal tool vs public API)