NYC

pathml

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill documentation includes instructions to install the 'pathml' package via 'uv pip install'. Since this package does not originate from a recognized trusted repository or organization, it is classified as an unverifiable dependency.
  • COMMAND_EXECUTION (LOW): The skill contains shell command snippets intended for package installation and environment setup.
  • PROMPT_INJECTION (LOW): Indirect Prompt Injection surface (Category 8) detected through the processing of diverse whole-slide image formats. 1. Ingestion points: 'SlideData.from_slide()' in SKILL.md. 2. Boundary markers: Absent from the processing logic. 3. Capability inventory: Local file system writes (HDF5 storage) and ML model inference. 4. Sanitization: No sanitization of image metadata or headers is specified, which could allow maliciously crafted files to influence agent behavior.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 03:44 AM