log-redaction-auditor

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly requires emitting a JSON report containing findings with a match field and examples drawn from the input logs, which means the agent would need to include matched secret values verbatim in its output (even if directed to write to a file), creating an exfiltration risk.