subagent-driven-development

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt explicitly instructs dispatching subagents to "read this first — it is your requirements, with the exact values to use verbatim" (i.e., include exact/magic strings from the brief in generated output), which forces any sensitive tokens placed in those briefs to be handled and emitted verbatim by the LLM, creating an exfiltration risk.