CVE-2025-15379 is a critical command injection vulnerability in MLflow versions prior to 3.8.1. It occurs in the model serving container initialization code when deploying models with the env_manager=LOCAL option. The vulnerability arises because MLflow reads dependency specifications from the model artifact's python_env.yaml file and directly interpolates them into a shell command without proper sanitization. This allows an attacker who can supply a malicious model artifact to execute arbitrary commands on the system with the same privileges as the MLflow process. The issue is fixed in MLflow version 3.8.1.