ai-avatar-video
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill refers to an installation script hosted on the vendor's official GitHub repository at
https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.mdfor thebeltCLI. - [COMMAND_EXECUTION]: The skill instructs the agent to use the
beltCLI for logging in and running AI model applications, which involves executing shell commands in the local environment. - [PROMPT_INJECTION]: Indirect Prompt Injection Surface:
- Ingestion points: The skill accepts external URLs for images, audio, and video files which are then processed by various AI models via the CLI (SKILL.md).
- Boundary markers: JSON inputs in shell commands are wrapped in single quotes in the provided examples.
- Capability inventory: Executes shell commands using the
belttool, including operations that write output to local files (SKILL.md). - Sanitization: There are no explicit instructions for validating or sanitizing the content of the remote files before they are processed.
Audit Metadata