lazagne
Fail
Audited by Socket on Sep 15, 2026
1 alert found:
MalwareMalwareSKILL.md
HIGHMalwareHIGH
SKILL.md
MALICIOUS. The skill is built around credential harvesting and includes an explicit remote download-and-execute example from an attacker host over HTTP. Even though LaZagne is a real upstream tool, this skill’s footprint and usage guidance are fundamentally incompatible with a benign agent skill.
Confidence: 96%Severity: 99%
Audit Metadata