chinese-app-benchmark
Warn
Audited by Socket on Apr 12, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the core purpose is coherent for UX benchmarking, but the skill overreaches by combining untrusted web research with autonomous code edits and external actions, and it explicitly bypasses git pre-push protections. No strong evidence of malware or credential theft appears, yet the execution model is high-trust and insufficiently scoped.
Confidence: 87%Severity: 69%
Audit Metadata