hunt-agent-deserialization

Warn

Audited by Socket on Nov 2, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

The skill is a penetration-testing guide with substantial exploit capability, including RCE and XXE payloads and a runnable probe. It does not show confirmed malicious intent, hidden execution, or credential theft, but its offensive techniques create high security risk if used against unauthorized targets.

Confidence: 96%Severity: 82%
Audit Metadata
Analyzed At
Nov 2, 2026, 10:16 PM
Package URL
pkg:socket/skills-sh/elementalsouls%2Fclaude-agenthunter%2Fhunt-agent-deserialization%2F@82d88e70ac4645dd6a6ea774cb42e61b1bd5227edcf2fde3c326e85fc1f9f015