designkit-ecommerce-product-kit
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local bash script (
run_ecommerce_kit.sh) to perform style creation, status polling, and image rendering. Parameters including product descriptions and image metadata are passed to this script via structured JSON input. - [EXTERNAL_DOWNLOADS]: Upon completion of the image generation process, the skill automatically downloads result files to the local filesystem. It uses a prioritized list of target directories, including the user's Downloads folder and workspace-specific paths.
- [PROMPT_INJECTION]: The skill's workflow creates an attack surface for indirect prompt injection through its data ingestion and processing steps.
- Ingestion points: The skill ingests user-supplied images (via URL or path) and text-based product descriptions to generate marketing selling points.
- Boundary markers: There are no explicit delimiters or instructions defined in the skill to distinguish between system instructions and embedded content within the processed product data.
- Capability inventory: The skill utilizes shell execution capabilities via
bashto run its internal design kit scripts. - Sanitization: The instructions do not describe any mechanisms for validating, escaping, or sanitizing the input data before it is interpolated into the JSON payload for script execution.
Audit Metadata