ecommerce-amazon-search-by-image
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides Python scripts (
amazon_search_by_image.pyandupload_image.py) designed to be executed by the agent to interface with Amazon search services and the vendor's asset management system. - [EXTERNAL_DOWNLOADS]: The skill fetches product metadata, including titles, brands, and sales data (via Keepa enrichment), from Amazon marketplaces through the vendor's proxy API.
- [DATA_EXFILTRATION]: The
upload_image.pyscript is capable of reading local image files or processing base64 data to upload them to the vendor's asset API. This is the intended mechanism for enabling visual searches of local content. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, presenting an attack surface for indirect prompt injection.
- Ingestion points: Ingests
imageUrlparameters and product descriptions/metadata returned from Amazon marketplaces. - Boundary markers: No explicit boundary markers or "ignore" instructions are defined for processing external product data.
- Capability inventory: The skill has the capability to perform network requests and read local files via the provided scripts.
- Sanitization: No explicit sanitization or filtering of external product titles or descriptions is described in the display rules.
Audit Metadata