amina-de-novo-protein-binder-design

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill defines a legitimate computational biology workflow for protein engineering. It utilizes recognized scientific models such as RFdiffusion, ProteinMPNN, and ESM-IF1 for their intended purposes.
  • [SAFE]: A human-in-the-loop mechanism is explicitly defined, requiring user confirmation of the design plan before proceeding to the computationally expensive backbone generation phase.
  • [SAFE]: No suspicious network operations, credential handling, or unauthorized file access patterns were identified in the provided instructions or reference files.
  • [NO_CODE]: The skill package consists entirely of documentation and parameter references in Markdown format and does not include any executable scripts or source code for direct audit.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 04:16 PM