linkfox-amazon-search-by-image

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The skill includes a utility scripts/upload_image.py that reads local files and uploads them to tool-gateway.linkfox.com. This is a documented functional requirement to convert local images into publicly accessible URLs for the Amazon visual search API. The destination belongs to the skill vendor.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data (product titles, brands, and descriptions) retrieved from Amazon listings. This is a standard ingestion surface for search-based skills.
  • Ingestion points: Product metadata returned by the search API in scripts/amazon_search_by_image.py.
  • Boundary markers: None identified; product data is displayed directly to the user.
  • Capability inventory: File system read access (via upload_image.py) and network communication (via urllib).
  • Sanitization: Not explicitly implemented in the provided scripts; the agent relies on its internal safety filters when rendering results.
  • [COMMAND_EXECUTION]: The skill documentation instructs the agent to execute local Python scripts (scripts/upload_image.py and scripts/amazon_search_by_image.py) to perform its tasks. These scripts use standard libraries and perform focused, legitimate operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 05:32 AM
Security Audit — agent-trust-hub — linkfox-amazon-search-by-image