sesame
Warn
Audited by Socket on Apr 23, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: The skill’s purpose matches Sesame integration, and the CLI source appears official via npm, so this is not confirmed malware. However, the actual data flow is through Membrane as an intermediary for authentication, token refresh, and proxy requests, which is disproportionate to a plain Sesame skill and creates meaningful third-party data and credential handling risk.
Confidence: 86%Severity: 64%
Audit Metadata