voice
Warn
Audited by Socket on Oct 24, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core local speech features are coherent and mostly proportionate, but the optional llm-router/OpenAI-compatible routing materially changes the trust boundary by sending speech data and possibly API credentials through third-party infrastructure. No confirmed malware or covert behavior is shown, yet the remote-backend and intermediary data-flow options make this higher risk than a purely local voice skill.
Confidence: 85%Severity: 64%
Audit Metadata