review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The review skill instructs the agent to inspect code, configs, and logs and to include concrete "Observed" evidence (exact file paths, line numbers, and brief proof), which can force the LLM to echo hard-coded secrets or credential values verbatim from reviewed files, creating an exfiltration risk.