theharvester

Warn

Audited by Socket on Sep 3, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS in overall use context because it gives an AI agent offensive security recon capability, but the skill content itself is internally coherent and shows no hidden exfiltration, malicious installer, or deceptive credential routing. Main risk comes from enabling pentest OSINT actions, not from evidence of malware.

Confidence: 93%Severity: 74%
Audit Metadata
Analyzed At
Sep 3, 2026, 02:39 AM
Package URL
pkg:socket/skills-sh/timsonner%2Fagent-skills%2Ftheharvester%2F@4568a6c0c7203572c6a683e2386a7022739622b091636535118782677b391fbe
Security Audit — socket — theharvester