cc-claude
Warn
Audited by Socket on Jun 19, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill's purpose is coherent, and the default DeepSeek endpoint is officially compatible, but the execution path relies on an unverified local shell wrapper and exposes API keys by collecting them in-chat and forwarding them on the command line. This looks more like risky credential handling and moderate supply-chain trust issues than confirmed malware.
Confidence: 85%Severity: 69%
Audit Metadata