bash-automation

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). Contains explicit reverse shell commands (bash, netcat, python) plus post‑exploitation enumeration (SUID, writable dirs) and remote command patterns that enable backdoors, remote code execution, and privilege escalation—high risk.

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

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill explicitly contains post-exploitation content and reverse-shell one-liners plus aggressive local enumeration (e.g., find /, SUID/writable scans) which enable remote control and compromise of the machine the agent runs on.