consult-oracle
Warn
Audited by Socket on Oct 23, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose is coherent, and the remote destination is the official ChatGPT domain, but the skill's core behavior relies on an unverifiable local automation script that can transmit selected project files through a signed-in browser session. That combination creates meaningful supply-chain and data exposure risk without clear provenance or credential-handling guarantees.
Confidence: 87%Severity: 72%
Audit Metadata