kaggle
Warn
Audited by Socket on May 7, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill is mostly coherent with a Kaggle integration and routes data to official Kaggle endpoints, but it has a broad action surface with real account side effects, untrusted-content processing plus Bash capability, and optional third-party skill-install channels that expand trust beyond Kaggle or the repo itself. No strong evidence of credential theft or covert exfiltration was shown.
Confidence: 86%Severity: 63%
Audit Metadata