gpt-image-2
Warn
Audited by Socket on Jul 28, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core capability fits an image-generation skill, and there is no malware-like installer or hidden execution path. However, the skill routes credentials and uploaded local images through AceDataCloud rather than OpenAI’s official API, and the public docs evidence does not fully match the skill’s claims about model/version and response behavior.
Confidence: 88%Severity: 62%
Audit Metadata