NYC
skills/smithery/ai/linux-apps/Gen Agent Trust Hub

linux-apps

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • COMMAND_EXECUTION (HIGH): The skill provides a wide range of shell commands to control the operating system's GUI. It can kill processes (pkill), close windows (wmctrl -c), and modify window states, which can be used to disrupt user activity or manipulate application behavior.
  • DATA_EXFILTRATION (HIGH): Functional commands for taking screenshots of the active window or specific applications are included (scrot, import). This poses a high risk of sensitive data exposure (e.g., capturing passwords, private messages, or confidential documents visible on screen).
  • EXTERNAL_DOWNLOADS (MEDIUM): The setup instructions require the use of sudo apt-get install to fetch multiple system utilities. While the sources are standard repositories, the requirement for root privileges during setup is a security concern.
  • Indirect Prompt Injection (HIGH): Ingestion points: Window titles and process lists are retrieved from the system environment. Boundary markers: None identified. The skill interpolates window titles directly into shell commands. Capability inventory: Screen capture, process termination, and window control. Sanitization: None; window titles are controlled by external applications (e.g., via a malicious website title) and are inherently untrusted, posing a risk of system manipulation.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 08:53 AM