offensive-wpa3-sae

Fail

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPRIVILEGE_ESCALATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to download the 'dragonblood' security toolset from a third-party GitHub repository (https://github.com/vanhoefm/dragonblood).
  • [REMOTE_CODE_EXECUTION]: Immediately after cloning the repository, the skill directs the execution of Python scripts (dragontime.py, dragondrain.py) from that external source, creating a direct path from download to code execution.
  • [COMMAND_EXECUTION]: The skill utilizes several wireless monitoring and attack tools, including airodump-ng, airbase-ng, and mdk4, to interact with network interfaces.
  • [PRIVILEGE_ESCALATION]: Multiple commands, including network sniffing and frame injection, require administrative rights and are prefixed with sudo, facilitating high-privilege operations on the host system.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 15, 2026, 01:30 PM
Security Audit — agent-trust-hub — offensive-wpa3-sae