baoyu-article-illustrator

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill employs shell commands (test -f) within its workflow to verify the existence of user preferences and reference images. These operations are limited to status checking and do not pose a risk of arbitrary command execution.
  • [PROMPT_INJECTION]: The image generation instructions include a directive to provide stylistically similar alternatives if sensitive figures are detected, rather than refusing. This is a common technique to mitigate over-refusal in creative tasks and does not constitute a malicious override of safety protocols.
  • [DATA_INGESTION]: The skill processes untrusted article content to extract labels and metrics for inclusion in generated illustrations. This data is interpolated into structured templates (ZONES/LABELS), which provides natural boundaries for the external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 05:59 PM