skills/advaitpaliwal/feynman/borzoi/Gen Agent Trust Hub

borzoi

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting and processing external genomic data and scientific literature, which creates an attack surface for indirect prompt injection.
  • Ingestion points: FASTA and variant manifests from source URLs or accession IDs, and comparative data from GTEx, ENCODE, or literature (SKILL.md).
  • Boundary markers: The skill does not define specific delimiters or instructions to ignore potential commands within the external datasets.
  • Capability inventory: The agent is instructed to save prediction arrays, summary tables, and plots to the filesystem as artifacts (SKILL.md).
  • Sanitization: There are no instructions for sanitizing or validating the content of the external files or literature before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 01:47 AM
Security Audit — agent-trust-hub — borzoi