watsonx-ai
Warn
Audited by Socket on Apr 28, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: The skill is internally coherent as a Membrane-based WatsonX connector, but its true footprint is broader than a direct WatsonX integration because authentication, connection management, and action execution are all routed through Membrane. Install trust is moderate rather than severe because the CLI comes from npm and appears same-vendor, but the third-party mediation of credentials and data makes this higher risk than a normal direct API skill.
Confidence: 85%Severity: 58%
Audit Metadata