mainframe-security
Audited by Socket on Sep 19, 2026
2 alerts found:
Securityx2SUSPICIOUS. The skill is coherent with its stated purpose, but that purpose is an AI-agent mainframe pentesting playbook with offensive, privilege-escalation, and stealth guidance, which is inherently high risk. Install trust is mixed: official-looking Hercules source build is plausible, but TK4- is fetched from an unpinned mutable GitHub archive and SEAR examples use disabled TLS checks. No clear credential-harvesting or exfiltration to unrelated third parties is shown, so this is not confirmed malware, but it is a high-risk offensive-security skill with notable supply-chain and execution concerns.
The excerpt is security-assessment and z/OS administration documentation. It contains no clear malware, persistence, credential theft, or covert exfiltration behavior. It does present substantial operational risk: TLS verification is disabled, sensitive RACF data is exported locally, credentials are handled insecurely, and highly privileged account, dataset, APF, and JCL operations are documented. Use only in authorized environments, enforce certificate validation, protect exports, avoid shell-visible passwords, and apply strict authorization and change controls.