katana-web-crawling
Warn
Audited by Socket on Apr 15, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill is internally consistent and uses official same-org install sources, so it does not look malicious or like credential harvesting. However, it equips an AI agent with web reconnaissance/crawling capability against external systems, which is an offensive-security function and therefore carries elevated security risk despite the explicit authorization warnings.
Confidence: 90%Severity: 72%
Audit Metadata