mom

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the agent to quote and reproduce user-provided passages verbatim (e.g., the "[Passage]" sections and "The part she'd actually remember"), so if the input text contains API keys, tokens, passwords, or other secrets the model will likely echo them in its output, creating an exfiltration risk.