MLFLOW · CVE-2025-15379 · Package
FIX → 3.8.1
MLflow builds a shell command from dependency entries in a model's python_env.yaml without sanitizing them when deploying with env_manager=LOCAL. A malicious model artifact can make the server run arbitrary commands on any system that deploys it.
first public Mar 30, 2026 · 1 public exploit · 2 sources
Check it with your coding agent
It reads this page and tells you if your code or app is affected.
Teams that deploy MLflow models with env_manager=LOCAL, especially when the model artifacts come from untrusted or shared sources. Deployments that only serve models you built and trust, or that use a different environment manager, are less exposed. The sources only name version 3.8.0 as affected and 3.8.1 as fixed, so other versions below 3.8.1 should be treated with care.
Not affected if default-config, trusted-artifacts-only
Who cares ml-platform, mlflow, model-serving, python
pip show mlflow, looking for 3.8.0.env_manager=LOCAL and load artifacts from sources you do not fully control.A command injection vulnerability exists in MLflow's model serving container initialization code, specifically in the _install_model_dependencies_to_env() function. When deploying a model with env_manager=LOCAL, MLflow reads dependency specifications from the model artifact's python_env.yaml file and directly interpolates them into a shell command without sanitization. This allows an attacker to supply a malicious model artifact and achieve arbitrary command execution on systems that deploy the model. The vulnerability affects versions 3.8.0 and is fixed in version 3.8.1.
Are you affected?
Ask from here. InfraStack checks the affected versions against your stack and answers in your inbox.
InfraStack checks the affected versions against your stack and answers in Slack, with the evidence and the fix.
InfraStack checks it against your stack and answers on your report, with the evidence and the fix.
Upgrade to 3.8.1
mlflow upgrade to 3.8.1GitHub Advisory Database, CC-BY 4.0 · NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.