imaging-data

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests NIfTI medical imaging data using nibabel and numpy in scripts/profile_imaging_dataset.py. This creates an attack surface for indirect injection if processed binary files contain malicious payloads. However, the skill functions as a passive auditor, extracting specific voxel statistics and metadata into structured JSON reports, which avoids the interpolation of untrusted data into dangerous sinks.
  • [EXTERNAL_DOWNLOADS]: The skill depends on the nibabel and numpy Python packages. These are well-known, legitimate scientific libraries widely used in the medical imaging community, and their inclusion is standard practice for the skill's primary objective of dataset profiling.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 01:27 PM