image-generation
Warn
Audited by Socket on Oct 16, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill’s capabilities fit image generation, and there is no evident hidden exfiltration or command execution. However, trust in the z-ai ecosystem is only partially verifiable: package/CLI provenance is ambiguous, versions are unpinned, and SDK docs allow configurable API endpoints, creating moderate supply-chain and data-routing risk.
Confidence: 84%Severity: 52%
Audit Metadata