skills/wubin1836/ai-ecommerce-expert-agent-skills/ai-ecommerce-expert-product-image-to-video/Gen Agent Trust Hub
ai-ecommerce-expert-product-image-to-video
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill dynamically downloads the
@infimind/ecom-content-clipackage from the npm registry usingnpx -yat runtime. The use of the@latesttag creates a dependency on the external registry's current state and could lead to unexpected behavior if the package is updated. - [COMMAND_EXECUTION]: The script
scripts/imiva_mcp.pyutilizessubprocess.Popento execute shell commands (npx). While the command is scoped to the specific IMIVA package, it represents a bridge between the AI environment and the host system's shell. - [REMOTE_CODE_EXECUTION]: By invoking
npx, the skill executes code retrieved from a remote source. This is the intended delivery mechanism for the tool but constitutes remote code execution as part of the skill's operation. - [DATA_EXPOSURE]: The skill is designed to handle local file paths for product images (
referenceImages). While this is a core feature, it grants the agent visibility into the local filesystem for specific files passed to the tool. - [INDIRECT_PROMPT_INJECTION]: As the skill processes external product data and user descriptions to generate video parameters, it possesses an attack surface for indirect prompt injection. The skill includes instructions for the agent to verify facts and goals as a mitigation strategy.
Audit Metadata