Posit Workbench container images for specialized environments

Container images that package Workbench for specific managed cloud platforms. Each image arrives configured for the host platform's authentication, storage, and lifecycle conventions, so you can use it directly as the workstation or compute image.

For the standard, non-platform-specific Workbench container images, see posit-dev/images-workbench.

Images

Image	Platform	Platform documentation
workbench-for-google-cloud-workstations	Google Cloud Workstations	Develop code using Posit Workbench
workbench-for-microsoft-azure-ml	Azure Machine Learning compute instances	Add custom applications such as RStudio or Posit Workbench

See each image's documentation for the canonical registry, available tags, and platform-specific configuration.

Prerequisites

Tool	Required for	Install
Container build tool	Building images locally	Docker, Podman, or Buildah
Product license	Running Workbench	Licensing FAQ

To consume the published images on their target platforms, follow the platform documentation linked above. You do not need any local container tooling.

Building from source

You can interact with this repository in multiple ways:

• Build container images directly from the Containerfile.
• Use the bakery CLI to manage and build container images.

Build

You can build Open Container Initiative (OCI) container images from the definitions in this repository using one of the following container build tools:

• buildah
• docker buildx
• podman

Each Containerfile uses the root of the repository as its build context. The bakery.yaml project file is in the root of this repository.

PWB_VERSION="2026.01"

# Build the Google Cloud Workstations standard image using docker
docker buildx build \
    --tag workbench-for-google-cloud-workstations:${PWB_VERSION} \
    --file workbench-for-google-cloud-workstations/${PWB_VERSION}/Containerfile.ubuntu2404.std \
    .

# Build the Azure ML standard image using buildah
buildah build \
    --tag workbench-for-microsoft-azure-ml:${PWB_VERSION} \
    --file workbench-for-microsoft-azure-ml/${PWB_VERSION}/Containerfile.ubuntu2404.std \
    .

# Build the Azure ML minimal image using podman
podman build \
    --tag workbench-for-microsoft-azure-ml:${PWB_VERSION} \
    --file workbench-for-microsoft-azure-ml/${PWB_VERSION}/Containerfile.ubuntu2404.min \
    .

Using bakery

This repository follows the structure described in bakery usage.

The following additional documentation is available:

• Configuration Reference: bakery.yaml schema and options
• Templating Reference: Jinja2 macros for Containerfile templates
• CI Workflows: shared GitHub Actions workflows for building and pushing images

Prerequisites

The following build prerequisites are required:

• python
• uv
• docker buildx bake
• gh (required while repositories are private)
• bakery
• goss and dgoss for running image validation tests

Build with bakery

By default, bakery creates an ephemeral JSON bakefile to render all containers in parallel.

bakery build

You can view the bake plan using bakery build --plan. Use the CLI flags to build only a subset of images in the project.

Test images

After building the container images, run the test suite for all images:

bakery run dgoss

You can use CLI flags to limit the tests to a subset of images.

Related repositories

This repository is part of the Posit Container Images ecosystem. The non-specialized Workbench images live in images-workbench. For shared build tooling and CI workflows, see images-shared.

Share your feedback

We invite you to join us on GitHub Discussions to ask questions and share feedback.

Issues

If you encounter any issues or have any questions, please open an issue. We appreciate your feedback.

Code of Conduct

We expect all contributors to adhere to the project's Code of Conduct and create a positive and inclusive community.

License

Posit licenses these container images and associated tooling under the MIT License.