ai-avatar-video

Pass

Audited by Gen Agent Trust Hub on Nov 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing a CLI tool using npx skills add belt-sh/cli and references installation instructions from a GitHub repository (raw.githubusercontent.com/inference-sh/skills). These are recognized vendor resources for the platform described.
  • [COMMAND_EXECUTION]: The skill provides numerous examples using the belt CLI tool to run various AI models (e.g., pruna/p-video-avatar, bytedance/omnihuman-1-5). These commands are standard operations for the tool's intended purpose and involve passing parameters for image/audio URLs and text scripts.
  • [DATA_EXPOSURE_EXFILTRATION]: While the skill uses external URLs for images and audio files (e.g., https://portrait.jpg, https://speech.mp3), these are provided as placeholders for user-supplied data and do not represent exfiltration of sensitive local information.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided text scripts for text-to-speech generation. While this is an ingestion point for untrusted data, the impact is limited to the content of the generated audio and does not grant the skill unauthorized capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 18, 2026, 08:23 PM