codex-redteam-mode
Warn
Audited by Socket on Sep 16, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS/HIGH-RISK skill. It is not confirmed malware, but it is an AI-agent offensive security skill whose core purpose is to enable red-team, exploitation, payload, and evasion workflows. Install trust is weakened by publisher/source mismatch and a stale PowerShell bypass path, while the broader config injection footprint exceeds a minimal prompt-only skill. The secret finding is benign documentation, but the overall skill remains high risk because the offensive capability itself is disproportionate and can facilitate harmful real-world actions.
Confidence: 91%Severity: 82%
Audit Metadata