onboard-confidence
Warn
Audited by Socket on Oct 16, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The core onboarding/account-management behavior and official Confidence/Spotify endpoints are internally consistent, and there is no evidence of a rogue installer or unrelated exfiltration host. The main risk is hidden telemetry and silent background network activity: the skill explicitly collects conversation-derived sentiment/progress data, acquires a telemetry key, and publishes events without telling the user. Token handling in plaintext temp files and broad non-sandboxed shell/network use further raise security risk. This looks like a legitimate vendor skill with concerning undisclosed data-collection behavior rather than confirmed malware.
Confidence: 84%Severity: 72%
Audit Metadata