view-images

Pass

Audited by Gen Agent Trust Hub on Oct 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data from files, screen captures, and web pages, creating a surface for indirect prompt injection where malicious instructions could be hidden in visual or document content.
  • Ingestion points: SKILL.md describes workflows using nolune view, screencapture, and Google Chrome --screenshot to pull data from the filesystem, the active screen, and external URLs.
  • Boundary markers: No explicit delimiters or instructions are used to warn the agent about potentially malicious instructions embedded within the processed media.
  • Capability inventory: The skill uses nolune view to send data to the conversation context and leverages multiple CLI tools (sips, magick, ffmpeg, qlmanage, pdftoppm) for file conversion and manipulation.
  • Sanitization: The skill does not implement sanitization or filtering for the visual content it processes before passing it to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 30, 2026, 07:17 PM