codex-autoresearch-loop
Warn
Audited by Socket on Sep 12, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill's core purpose matches its code-editing and git capabilities, but its footprint is riskier than typical dev workflow skills: third-party skill installation from a personal repo despite different branding, autonomous unbounded execution, and web-search-driven iteration with write/exec powers. No confirmed malware or credential theft is shown, but the trust chain and unattended autonomy make it high security risk.
Confidence: 90%Severity: 81%
Audit Metadata