manuscript-review-revise
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were identified. The skill follows best practices for a prompt-based editing tool.
- [DATA_EXFILTRATION]: While the skill creates output files in a 'reports/' directory, it does not attempt to send data to external servers or access sensitive user credentials.
- [PROMPT_INJECTION]: The skill processes user manuscripts which are external data. It manages the risk by setting strong internal constraints that prevent the modification of scientific conclusions or the addition of unverified data. No specific delimiters are used, but the task context is highly constrained.
Audit Metadata