cloud-pentest
Warn
Audited by Socket on May 9, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill is internally coherent for a cloud security assessment, but it equips an AI agent with offensive cloud reconnaissance and exploitation-adjacent capability using powerful cloud credentials. Risk is driven mainly by offensive-security enablement and unpinned third-party tool installation, not by clear evidence of credential theft or malicious data exfiltration.
Confidence: 89%Severity: 78%
Audit Metadata