car-advisor
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external automotive media and forums, which is a potential entry point for malicious instructions.
- Ingestion points: Content is retrieved via web searches and direct access to domains like dongchedi.com, autohome.com.cn, zhihu.com, and various car manufacturer websites as listed in SKILL.md and references/data-sources.md.
- Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to ignore or escape commands potentially embedded in external review content.
- Capability inventory: The skill focuses on information retrieval and presentation; no dangerous file system, shell execution, or privilege escalation capabilities were detected.
- Sanitization: There are no defined procedures for sanitizing or validating the data fetched from external sources before it is processed by the agent.
Audit Metadata