image-to-video
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
beltCLI tool to perform AI inference tasks. It provides examples of executing commands likebelt app runto process images and generate video content. These operations are restricted to thebeltcommand as specified in the skill's configuration. - [EXTERNAL_DOWNLOADS]: The skill references external resources for installation and model interaction, specifically targeting the
inference-shGitHub repository and theinference.shdomain. These downloads are legitimate and necessary for the skill's primary function of providing access to video generation models. - [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected. The skill follows best practices for secret management by instructing the user to use
belt loginrather than hardcoding credentials, and it uses structured JSON for model inputs to minimize injection risks.
Audit Metadata