What if we measure how bad code is using AI and math?

GitHub - sebastiancarlos/snafu: SNAFU. Agentic LLM flow to help you with "naming things" in source code.