codex
Warn
Audited by Socket on Oct 8, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose is coherent with coding delegation, and the upstream Codex CLI appears official, but the skill’s actual trust boundary is a local bundled shell script that users are told not to inspect. That black-box wrapper, combined with full-workspace autonomous access and unspecified outbound data flow, makes the skill higher risk than a normal official CLI integration.
Confidence: 86%Severity: 81%
Audit Metadata