screenshot

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Python, PowerShell, and Swift to invoke system-level screenshot tools such as screencapture (macOS) and scrot (Linux). These commands are executed locally using scripts included in the skill package and are scoped to the intended functionality.
  • [DATA_EXFILTRATION]: While the skill accesses the user's screen content, it exclusively saves the resulting image files to the local file system. There is no code present for network transmission or data exfiltration.
  • [SAFE]: The skill does not exhibit patterns of prompt injection, persistence, or obfuscation. It functions as a standard utility for enhancing the agent's environmental awareness through visual capture.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 12:23 PM
Security Audit — agent-trust-hub — screenshot