VLM_Expert

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [SAFE] (SAFE): The file contains only markdown documentation and usage examples. No executable scripts, system-level commands, or remote dependencies were identified.
  • [Indirect Prompt Injection] (LOW): As a vision interface, the tool ingests external image data which is an inherent surface for indirect prompt injection; however, no specific vulnerabilities or side-effect capabilities are present in the skill definition.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 11:59 AM