god-syndicate

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs exfiltration and to report "findings and extracted data" and "IOCs", which would require the model to include secret/credential values verbatim in its output.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is explicitly an offensive toolkit/instruction set that describes and facilitates malicious actions (C2/RAT development, exfiltration, credential theft, persistence, and extensive evasion techniques).

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs privileged and state-changing offensive actions (privilege escalation, kernel exploits, EDR/AV/OS patching, log wiping, C2/RAT deployment and fileless payloads) that would compromise the host system.