requirements-doc

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt explicitly instructs the agent to "capture everything the user says — don't filter out ideas prematurely" and save the unfiltered stream-of-consciousness to a file, which would cause any user-provided secrets to be echoed verbatim in the generated document and stored, creating an exfiltration risk.