modality-detection

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements logic for parsing clinical text and DICOM metadata to classify imaging studies (e.g., CT, MRI, X-ray).
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text and DICOM files as input. However, the logic is restricted to non-executable keyword matching (e.g., .upper() and in checks) and standard attribute access using the pydicom library. There are no capabilities for command execution, file writing, or network operations that could be exploited via malicious input.
  • [EXTERNAL_DOWNLOADS]: The skill references the pydicom Python package, which is the standard, well-known library for DICOM processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 11:48 AM