tooluniverse-protein-therapeutic-design
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its reliance on external data sources for target characterization.
- Ingestion points: Data is fetched from UniProt, PDB, and EMDB via tools like
tu.tools.UniProt_get_protein_by_accessionandtu.tools.PDB_get_structurein SKILL.md. - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the fetched biological data.
- Capability inventory: The skill performs complex structural modeling using NVIDIA NIM APIs and generates report files (Markdown, FASTA, CSV) based on the input.
- Sanitization: No evidence of sanitization or structural validation is present for data retrieved from public bioinformatics databases.
- [EXTERNAL_DOWNLOADS]: The skill retrieves protein structures and sequences from established scientific services including the Protein Data Bank (PDB), UniProt, and the Electron Microscopy Data Bank (EMDB). It also makes network calls to official NVIDIA NIM API endpoints for protein backbone generation and sequence design. These sources are considered well-known and trusted in the technology and scientific communities.
Audit Metadata