code-review
Warn
Audited by Socket on Aug 16, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core code-review behavior is coherent and proportionate, but the skill introduces medium risk through transitive installation of `/setup-utarn-skills` and processing of untrusted external/repo content in sub-agents. No direct credential harvesting, covert behavior, or clear malicious data exfiltration is present in the provided skill text.
Confidence: 84%Severity: 52%
Audit Metadata