refactor

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill requires producing verbatim "before/after" code to preserve behavior and gives no guidance to redact or replace embedded credentials, so if the user's code contains API keys/tokens/passwords the model would likely reproduce them, creating a high exfiltration risk.