analytics
Warn
Audited by Socket on Sep 20, 2026
1 alert found:
AnomalyAnomalyagents/openai.yaml
LOWAnomalyLOW
agents/openai.yaml
No direct malicious behavior or obfuscated payload is present in this configuration. It does create a meaningful potential security risk by giving an analytics agent arbitrary command execution and file modification capabilities. Review tool sandboxing, command allowlists, file-scope restrictions, and the contents of SKILL.md before deployment.
Confidence: 96%Severity: 58%
Audit Metadata