conversation-lab

Warn

Audited by Socket on Mar 17, 2026

1 alert found:

Anomaly
AnomalyLOW
SKILL.md

SUSPICIOUS: The stated purpose matches conversation-analysis orchestration, and no explicit credential theft or external exfiltration is shown. However, the skill materially relies on unseen local scripts and multiple composed skills, creating a transitive trust boundary with unverifiable provenance and moderate autonomous behavior.

Confidence: 79%Severity: 61%
Audit Metadata
Analyzed At
Mar 17, 2026, 06:42 AM
Package URL
pkg:socket/skills-sh/grahama1970%2Fagent-skills%2Fconversation-lab%2F@8dadea0c4afea8af822b4bd209a6acffd00d6a77
Security Audit — socket — conversation-lab