ai-ecommerce-expert-smart-product-retouch

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The helper script scripts/imiva_mcp.py uses subprocess.Popen to invoke the @infimind/ecom-content-cli package. The execution is handled without a shell, and arguments are passed as a list, which is a secure implementation for the skill's core functionality.- [EXTERNAL_DOWNLOADS]: The skill uses npx to fetch the @infimind/ecom-content-cli tool. This package belongs to the official service provider (Beijing Infimind Technology Co., Ltd.) described in the documentation and is considered a safe dependency within the context of the service's ecosystem.- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing external data.
  • Ingestion points: The images and prompt arguments in scripts/imiva_mcp.py are populated by user-provided input and product data.
  • Boundary markers: The instructions guide the agent to maintain product facts, but no technical delimiters or sanitization are applied to the ingested strings.
  • Capability inventory: The skill uses subprocess calls to execute commands and performs network operations via the MCP CLI to transmit data to the service provider.
  • Sanitization: Content is passed directly into the JSON-RPC request without additional filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 08:02 AM
Security Audit — agent-trust-hub — ai-ecommerce-expert-smart-product-retouch