devirtualizing-vm-protected-code
Warn
Audited by Socket on Aug 7, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill is internally coherent as a reverse-engineering guide and shows no direct credential theft or exfiltration, but it equips an AI agent with high-risk protected-code devirtualization capability that can be used to defeat commercial software protections. Supply-chain risk is limited because no installer is embedded, though some referenced tooling has weaker provenance than official registry-distributed tools.
Confidence: 87%Severity: 72%
Audit Metadata