dicom-metadata-extract
Warn
Audited by Snyk on Jun 18, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.70). At runtime the skill reads the caller-provided DICOM file (
dicom_path) viapydicom.dcmread(..., stop_before_pixels=True)and ingests its header text fields (e.g., PatientName/PatientID) into the returned JSON that is then placed into the agent’s LLM context; if that DICOM is outsider-authored, it becomes outsider free text.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata