jellyreach
Warn
Audited by Socket on May 1, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The skill's capabilities mostly match its stated Jellyreach-management purpose, and the Membrane CLI install path appears same-vendor and officially documented. The main concern is data-flow integrity: instead of talking directly to Jellyreach's official API, the skill routes authentication and action execution through Membrane as an intermediary, concentrating trust and exposing Jellyreach data/credentials to a third-party platform. This is disclosed rather than hidden, so it is not confirmed malicious, but it is a meaningful security and trust expansion beyond a direct integration.
Confidence: 87%Severity: 58%
Audit Metadata