CacheDeception
Warn
Audited by Socket on Nov 16, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
The skill is coherently focused on web cache exploitation, with no evident credential exfiltration, malicious installer, or command execution. Its main concern is operational: it gives an AI agent detailed instructions for high-impact offensive testing, including retrieving victim data and poisoning shared caches, without explicit per-action approval or clear safeguards restricting tests to authorized, controlled accounts. Classify as suspicious/high risk for autonomous exploitation capability, not confirmed malware.
Confidence: 93%Severity: 76%
Audit Metadata