gpt-6
Warn
Audited by Socket on Sep 22, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill's core function is coherent as a model-proxy client, but it is not a direct OpenAI skill despite presenting itself that way. The main risk is intermediary routing of prompts and credentials through Frevana plus unverifiable behavior in the missing local script, not confirmed malware.
Confidence: 84%Severity: 66%
Audit Metadata