pydicom
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious code, obfuscation, or unauthorized access patterns were detected. The skill provides standard utilities for medical imaging workflows and uses well-known, legitimate dependencies.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external DICOM files, which represents an indirect prompt injection surface.
- Ingestion points: The pydicom.dcmread() function is used across SKILL.md and all scripts in the scripts/ directory to ingest external medical data.
- Boundary markers: Absent. There are no explicit delimiters or instructions to ignore embedded commands within the ingested DICOM metadata.
- Capability inventory: The skill can perform file system writes via Dataset.save_as() and Image.save(), though it lacks network or subprocess execution capabilities.
- Sanitization: Absent. Metadata from DICOM fields is processed and displayed without specific sanitization for natural language instructions.
Audit Metadata