pydicom

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process external medical imaging files (DICOM) which contain numerous user-controllable metadata fields. \n
  • Ingestion points: External data enters the agent context via pydicom.dcmread() in scripts/anonymize_dicom.py, scripts/dicom_to_image.py, and scripts/extract_metadata.py.\n
  • Boundary markers: The skill does not implement boundary markers or instructions to the agent to disregard potential commands embedded within the DICOM tags.\n
  • Capability inventory: The skill possesses the capability to write files to the local system using Dataset.save_as() in scripts/anonymize_dicom.py and Image.save() in scripts/dicom_to_image.py. It also outputs extracted metadata to the console or files via scripts/extract_metadata.py.\n
  • Sanitization: There is no evidence of sanitization or content validation for the strings extracted from DICOM tags before they are processed or displayed.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 08:48 PM