ecommerce-ebay-search
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests product data from eBay via a proxy API, which includes untrusted content generated by eBay sellers. This data is added to the agent's context without sanitization or explicit delimiters, potentially allowing an attacker-controlled listing to inject instructions.
- Ingestion points: The tool fetches listing data (titles, subtitles, prices) via the
scripts/ebay_search.pyscript and returns it to the agent. - Boundary markers: The instructions do not define clear boundaries or warnings for the agent to ignore instructions embedded within the retrieved product data.
- Capability inventory: The agent can execute the search script and read/write JSON data files within the project directory.
- Sanitization: The skill performs no filtering or sanitization of listing content before it is processed by the AI.
Audit Metadata