The Agent Skills Directory

COMMAND_EXECUTION (SAFE): The code_validator.py script uses subprocess.run with argument lists to execute ruff and mypy. This approach is safe as it avoids shell interpretation and prevents command injection through the path argument.
EXTERNAL_DOWNLOADS (SAFE): While async_fetcher.py is designed to download content from URLs, it implements protocol validation (requiring http:// or https://) and is intended for legitimate testing and demonstration of async patterns.
CREDENTIALS_UNSAFE (SAFE): config_loader.py demonstrates safe secret handling using Pydantic's SecretStr and includes logic to prevent the use of default development secrets in production environments.
DATA_EXFILTRATION (SAFE): No patterns of unauthorized file access or exfiltration of sensitive information (such as .ssh or .aws credentials) were detected in the source code.

mastering-python-skill