mlops-deployment

Warn

Audited by Socket on Sep 29, 2026

1 alert found:

Anomaly
AnomalyLOW
scripts/model_server.py

No clear malicious behavior is present. The main security risk is unsafe deserialization if the configured model file is untrusted, compounded by an unauthenticated reload endpoint. Publicly exposed endpoints, unbounded request sizes, and exception-detail disclosure also warrant review.

Confidence: 97%Severity: 64%
Audit Metadata
Analyzed At
Sep 29, 2026, 01:41 PM
Package URL
pkg:socket/skills-sh/pluginagentmarketplace%2Fcustom-plugin-ai-data-scientist%2Fmlops-deployment%2F@7e058e0d1138c808aa8ad763d9059284d5616bfcc31ef0e6ccfce73b9ef8f102