setup

Pass

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPRIVILEGE_ESCALATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches installation scripts and system tools from well-known technology services and trusted repositories, including Homebrew's official installation script (github.com/Homebrew/install) and NodeSource's repository configuration for Linux (nodesource.com).
  • [EXTERNAL_DOWNLOADS]: Downloads machine learning models (specifically Faster-Whisper, Yunet, and Robust Video Matting) required for video transcription and visual effects.
  • [COMMAND_EXECUTION]: Executes shell commands to perform environment diagnostics and install dependencies through established package managers like brew, winget, apt, dnf, and pacman.
  • [PRIVILEGE_ESCALATION]: Uses sudo for system-wide tool installations on Linux. The instructions follow a safe pattern by advising the agent to prompt the user to execute these privileged commands manually in their own terminal.
  • [CREDENTIALS_SAFE]: Implements secure handling of API keys by reading them directly from the system clipboard into a local configuration file, avoiding exposure in the chat history. To mitigate injection risks, the script validates these keys against a restrictive regular expression (^[A-Za-z0-9_\-.]{20,}$) before storage. Sensitive files like aws.env are created with 0600 permissions (readable/writable only by the owner).
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 11, 2026, 01:29 PM