ai-ecommerce-expert-nano-banana-2-ecommerce-image

Warn

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The script scripts/imiva_mcp.py invokes npx -y @infimind/ecom-content-cli@latest via subprocess.Popen. This pattern downloads and executes code from a third-party source at runtime without a pinned version, allowing for unverified code updates to be executed automatically.
  • [COMMAND_EXECUTION]: The skill uses subprocess.Popen to execute system commands (npx) to manage the IMIVA Ecommerce MCP server lifecycle.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs which are then used as parameters for the IMIVA API, creating a surface for indirect injection attacks. * Ingestion points: The prompt and images arguments in the create_smart_refine_task tool defined in scripts/imiva_mcp.py. * Boundary markers: None identified; user input is passed directly to the tool arguments without isolation. * Capability inventory: The skill can perform image refinement, query task history, and check user credits via a remote service. * Sanitization: The script performs no validation or sanitization of input strings before they are sent to the external command.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 25, 2026, 08:02 AM
Security Audit — agent-trust-hub — ai-ecommerce-expert-nano-banana-2-ecommerce-image