Expanded capabilities for physical AI
Expanded Integration: GitHub Packages & AWS ECR Now Supported
Chassy now supports pulling artifacts directly from GitHub Packages and AWS Elastic Container Registry (ECR). More flexibility when building and deploying your containers..
An artifact is anything your Continuous Integration system produces, including compiled binaries, archives, containers, and more. With this release, you can now easily import artifacts from GitHub Packages and Amazon ECR.
Simply add an Import Package step in a workflow, paste your container name in the “Artifact Name” field, e.g.,
ECR - 1234567890.dkr.ecr.us-west-2.amazonaws.com/my-company/my-image:latest
GitHub Packages - ghcr.io/my-company/my-image:1.0
For the full list of supported sources, check out our Workflow Steps documentation.
Deployment Enhancements: Package Invocations
Some packages can be invoked in different ways. Sometimes you may want to provide a configuration file or specify a subcommand. Package Invocations enable a new level of flexibility with how packages are executed in your deployments.
In the example above, under advanced options, we invoke the same package multiple times with different arguments. You can learn more about Package Invocations from an example in our documentation.
Triggering GitHub Actions workflows with Chassy
A core tenet of Chassy's mission is to streamline the work of robotics engineers by consolidating disparate services into a unified platform. Introducing the new Run CI Job workflow step, users can now trigger GitHub Actions workflows directly from Chassy and seamlessly import all container artifacts generated during workflow execution.
Our tutorial outlines a specific example of running a GitHub workflow, importing its artifacts and creating a release. For further details on the specifics of the Run CI Job step refer to its documentation.
Executable Archives: Bundled Binaries, Configs, and More
Chassy now natively supports executable archives as first-class deployment artifacts, right alongside binaries and containers.
Executable archives are perfect for use cases where your application needs to bundle additional files like configuration data, static assets, or other dependencies. Think of them like Java’s JAR files, AWS Lambda zip or Bazel archives: a packaged unit that includes both your binary and the resources it needs to run.
Each archive defines an entrypoint, acting as the launcher for your application. In a workflow, you can run the same archive multiple ways, just pass different arguments for different modes of operation. For example, use a config file for one subcommand, and call another directly for a different behavior.
Check out our Executable Archives Tutorial for an in-depth walkthrough on deploying with this new artifact type.
We’d like to hear what you think! For feedback or questions, email us at support@chassy.io
