ecommerce-ozon-shop-search
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a Python script (
scripts/ozon_shop_search.py) to communicate with the Seerfar API. The script uses standard Python libraries for network requests and local file management within the project directory. - [INDIRECT_PROMPT_INJECTION]: The skill ingests product data from an external marketplace (Ozon), which constitutes a potential surface for indirect prompt injection if the data contains adversarial text.
- Ingestion points: The API response is retrieved by
scripts/ozon_shop_search.pyand presented to the agent. - Boundary markers: The skill does not utilize specific delimiters to isolate the external data from its own instructions.
- Capability inventory: The skill can make POST requests to the Seerfar API proxy and write data to the local
nexscope/directory for caching purposes. - Sanitization: The script performs JSON validation but does not filter the text content for potential injection strings.
Audit Metadata