From a DevOps process standpoint, it is best practice to keep changes to an application in version control. Which of the following will allow changes to a docker Image to be stored in a version…

The `docker commit` command creates a new image from a running container's state, but the changes are not recorded in a human-readable, version-controlled script, making it non-reproducible and an anti-pattern for version control.

The `docker save` command exports an image to a tar archive, which is a binary representation of the image and cannot be version-controlled in terms of its changes or definitions.

A `docker-compose.yml` file defines how multi-container applications are run and orchestrated, but it does not specify how individual Docker images are built or track their internal changes for version control.

A Dockerfile provides a scriptable, declarative definition of how an image is built, including its base image, dependencies, and application code. Placing the Dockerfile under version control allows all changes to the image's definition to be tracked, ensuring reproducibility and enabling a consistent CI/CD pipeline.