the-lancet
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a domain-specific assistant for medical manuscript evaluation. It defines editorial heuristics and submission requirements without executing code or performing sensitive operations.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided manuscripts, which constitutes an ingestion point for untrusted data. However, the skill does not invoke any dangerous capabilities such as network requests, file modifications, or system commands, mitigating the potential for exploitation.
- Ingestion points: User-provided clinical or global-health manuscripts.
- Boundary markers: None explicitly defined in the instructions.
- Capability inventory: The skill logic is limited to text analysis and formatting; no external tools or shell commands are requested in the skill body.
- Sanitization: Not specified.
Audit Metadata