skills/wubin1836/ai-ecommerce-expert-agent-skills/ai-ecommerce-expert-amazon-a-plus-ecommerce-content/Gen Agent Trust Hub
ai-ecommerce-expert-amazon-a-plus-ecommerce-content
Warn
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The script
scripts/imiva_mcp.pyusessubprocess.Popento runnpx -y @infimind/ecom-content-cli@latest. This downloads and executes the latest version of an external Node.js package at runtime without version pinning. This behavior allows for potential execution of arbitrary code if the package is compromised. - [COMMAND_EXECUTION]: The skill executes shell commands via a Python wrapper to communicate with the MCP server. This grants the agent the ability to spawn external processes on the host system.
- [EXTERNAL_DOWNLOADS]: The skill depends on external software hosted on the NPM registry and communicates with
imiva.ecpro.com. These dependencies are not from a pre-defined list of trusted vendors. - [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection (Category 8).
- Ingestion points: User-supplied product information and media paths are passed as command-line arguments to the
create_detail_page_tasktool viascripts/imiva_mcp.py. - Boundary markers: There are no explicit delimiters or instructions to isolate user-provided data from command parameters.
- Capability inventory: The skill can execute commands, read local files, and perform network requests via the external tool.
- Sanitization: The skill does not perform validation or sanitization of external input before it is used in a sub-process call.
Audit Metadata