oracle-samples/ml-model-dynscan

A security-focused framework for dynamically scanning and testing machine learning models in an isolated Docker environment, detecting suspicious behaviors such as unauthorized network access or system modifications.

22
/ 100
Experimental

This framework helps ML developers and security engineers detect suspicious behaviors in machine learning models before deployment. It takes an ML model and a custom test script as input, then runs the model in a secure, isolated Docker environment to identify unauthorized network access, file system modifications, or dangerous system calls. The output is a report detailing any potential security issues found, ensuring the integrity and safety of the models.

No commits in the last 6 months.

Use this if you need to automatically scan your machine learning models for malicious or risky behaviors in a controlled environment.

Not ideal if you're looking for a tool to evaluate model performance or accuracy rather than security vulnerabilities.

MLSecOps Model Security Software Supply Chain Container Security Dynamic Analysis
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 5 / 25
Maturity 15 / 25
Community 0 / 25

How are scores calculated?

Stars

11

Forks

Language

Python

License

UPL-1.0

Last pushed

Aug 05, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/oracle-samples/ml-model-dynscan"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.