hunt-agent-browser-use
Warn
Audited by Socket on Nov 2, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
This is an offensive security-testing guide with a working prompt-injection and data-capture probe. Its stated purpose and safeguards are coherent, and there is no evidence of hidden exfiltration by the skill itself; however, the techniques can expose authenticated-session data if used outside a controlled test environment.
Confidence: 94%Severity: 72%
Audit Metadata