ai-avatar-video
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: No malicious behavior, such as prompt injection, data exfiltration, or obfuscation, was detected. The skill primarily serves as an instructional guide for the inference.sh platform.
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool, specifically restricted to theinfshcommand. This allows the agent to interact with the inference.sh API to run AI models and manage user sessions (e.g.,infsh login,infsh app run). - [EXTERNAL_DOWNLOADS]: The skill mentions the installation of a related skill (
inference-sh/skills@agent-tools) using thenpx skills addcommand. This is a standard mechanism for extending agent functionality within the ecosystem and targets the platform's own repository.
Audit Metadata