kagi-usage
Warn
Audited by Socket on Sep 13, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The task scope is coherent for web/page/mail retrieval, but the trust model is not: this skill relies on an unofficial third-party CLI, may forward Kagi credentials and private mail data through it, and instructs transitive loading of other skills. The main concern is supply-chain and credential-handling risk rather than confirmed malicious behavior.
Confidence: 89%Severity: 81%
Audit Metadata