screenshot

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local system utilities including screencapture (macOS), scrot, gnome-screenshot, and import (Linux) to capture screen data. This is the core functionality of the skill.
  • [DYNAMIC_EXECUTION]: The Python helper dynamically generates and executes AppleScript (via osascript) to activate target applications and uses the swift interpreter to run helper scripts for window and permission management. These operations are restricted to local system interaction for the purpose of taking screenshots.
  • [PRIVILEGE_ESCALATION]: The Windows implementation uses PowerShell's -ExecutionPolicy Bypass to allow the execution of the bundled .ps1 script. While this bypasses default security policies, it is a common requirement for running local automated scripts in development environments.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests window titles and application names from the operating system. While this creates a potential surface for indirect injection into the agent's context, the skill implements sanitization (quote escaping) when using these strings in shell commands, and the behavior is required for the tool's primary purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 06:07 PM
Security Audit — agent-trust-hub — screenshot