medical-specialty-briefs
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is composed exclusively of documentation and instructions, with no accompanying executable code, shell scripts, or binary assets. All references to medical journals are for public, reputable sources.
- [NO_CODE]: There are no software dependencies, packages, or scripts included, which eliminates risk categories related to code execution and persistence.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for summarizing external data from medical journals. Although it lacks explicit data delimiters or sanitization logic, it is assessed as safe because it does not possess tools for file system access, network exfiltration, or command execution.\n
- Ingestion points: Research articles retrieved from priority journals like NEJM, The Lancet, and JAMA.\n
- Boundary markers: The instructions do not include specific delimiters to isolate external research text from the prompt instructions.\n
- Capability inventory: The skill does not include any scripts or tools that allow the agent to perform privileged system operations.\n
- Sanitization: No explicit sanitization or filtering of journal content is prescribed.
Audit Metadata