protein-function-prediction
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No prompt injection patterns or attempts to override agent behavior were detected in the instructions or code.- [SAFE]: The skill does not perform any unauthorized data access or exfiltration; network usage is limited to retrieving protein data from UniProt via established tools.- [SAFE]: No obfuscated code, hidden strings, or suspicious encoding techniques were found within the skill's files.- [SAFE]: The Python implementation uses the open_biomed library for its intended purpose of scientific inference, following standard practices for loading model checkpoints and processing sequences.
Audit Metadata