explain

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill reads arbitrary source files and generates detailed explanations (including variable names, line numbers and literal code excerpts) and writes them to docs and the conversation without any instruction to redact secrets, so any embedded API keys/passwords could be reproduced verbatim.