codex-review
Warn
Audited by Socket on Aug 11, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: the stated purpose is plausible, but this snippet mostly instructs the agent to install a separate community skill whose actual instructions, permissions, and data flows are not visible. The main risk is transitive trust and unpinned remote skill installation, not confirmed malware.
Confidence: 84%Severity: 72%
Audit Metadata