youtube-thumbnail-design
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the
beltCLI and additional related skills from theinference-shGitHub repository and platform. - [COMMAND_EXECUTION]: The skill uses the
beltcommand via Bash to interact with remote AI applications, manage logins, and run image generation models. - [INDIRECT_PROMPT_INJECTION]: The skill templates image generation prompts that interpolate user-defined content, creating a surface for potential instruction injection into the target AI models.
- Ingestion points: Prompt strings within the
belt app runcommands inSKILL.md. - Boundary markers: None present in the provided templates to delimit instructions from generated content.
- Capability inventory: Execution of the
beltCLI tool which performs network operations and remote application runs. - Sanitization: No explicit sanitization or filtering of prompt content is performed within the skill instructions.
Audit Metadata