scientific-immunoinformatics
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill employs standard bioinformatics libraries such as mhcflurry, Biopython, and anarci for legitimate scientific data processing tasks.
- [SAFE]: No suspicious network requests, data exfiltration patterns, or unauthorized file access attempts were detected. Interactions with external databases (IEDB, IMGT, SAbDab) are conducted via defined tools for research purposes.
- [SAFE]: The instructions and code samples are transparent, containing no obfuscation, prompt injection attempts, or persistence mechanisms.
- [SAFE]: Analysis of the biological sequence ingestion (Indirect Prompt Injection surface) shows that the data is processed using mathematical models and scientific algorithms without execution of untrusted commands or logic.
Audit Metadata