track-tariff-trends
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external trade data from the worldmonitor.app API, which is a vector for indirect injection.
- Ingestion points: API responses from worldmonitor.app (SKILL.md).
- Boundary markers: The skill contains a 'Content safety' section instructing the agent to ignore directive-like text in responses.
- Capability inventory: Capabilities are limited to data retrieval; no sensitive file or system access is requested.
- Sanitization: The skill provides instructions to the agent to treat all response fields strictly as data.
- [PROMPT_INJECTION]: The skill body contains examples of injection strings ('ignore previous instructions') within a safety context. These are intended to harden the agent against external data poisoning rather than to bypass agent safeguards.
Audit Metadata