improve-logging

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill instructs the agent to include "current" code snippets and log call contents in the report (e.g., "Current: ...") and to search for sensitive fields, so if the repository contains hard-coded secrets (API keys, tokens, passwords) the agent would likely output them verbatim unless the prompt mandates redaction — creating an exfiltration risk.