hunt-agent-auth-bypass
Warn
Audited by Socket on Nov 2, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
The skill is coherent as an authorization-bypass testing guide, but its runnable exploit probes enable offensive testing against real agent endpoints and transmit a supplied session cookie. Treat as high risk unless use is restricted to authorized targets. No evidence establishes credential theft, concealed behavior, or confirmed malware.
Confidence: 94%Severity: 75%
Audit Metadata