varg-mascot-jingle-ad
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill uses established media processing tools and well-documented AI models to perform its stated purpose. Analysis of the code and instructions reveals no evidence of malicious behavior, persistence, or data exfiltration.
- [INDIRECT_PROMPT_INJECTION]: The skill presents an ingestion surface for untrusted data. * Ingestion points: User-provided product descriptions and seller fact sheets used in prompt templates in sections 1 and 2. * Boundary markers: The skill does not use specific delimiters or instructions to prevent the model from following commands embedded within the product data. * Capability inventory: The agent has access to ffmpeg for media creation and curl for interacting with external APIs. * Sanitization: No data validation or sanitization is performed on user inputs before they are included in model prompts. This is documented as a potential attack surface but is considered low risk given the creative nature of the task.
Audit Metadata