jb-buying-judge
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions to define a specific agent persona. It does not include any executable scripts, binary files, or configurations for external tools.
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to read and process user-supplied answers. While this represents a data ingestion point, it does not pose a security risk in this context. 1. Ingestion points: User-provided answers to the Buying Judge questions. 2. Boundary markers: None. 3. Capability inventory: None (no tool access, network access, or shell commands). 4. Sanitization: None. Conclusion: The absence of any exploitable capabilities prevents the ingested data from being used for malicious purposes.
Audit Metadata