This is the documentation for version 0.39. For documentation on the latest version of kpt, please see


Frequently asked questions

Q: What’s with the name?

A: kpt was inspired by apt, but with a Kubernetes focus. We wanted to uphold the tradition of naming tools to start with k, and also be short enough that you don’t have to alias it. It is pronounced “kept”.

Q: What does kpt provide that git clone doesn’t

A: kpt enables out-of-the-box workflows that git clone does not such as: cloning and versioning git subdirectories, updating from upstream by performing structured merges on resources, programmatically editing configuration (rather than with an editor), etc

Q: How is kpt different from other solutions?

A: Rather than expressing configuration as code, kpt represents configuration packages as data, in particular as YAML or JSON objects adhering to the kubernetes resource model

Q: Why resource configuration as the artifact rather than templates or configuration DSLs?

A: As explained in Declarative application management in Kubernetes, using resource configuration provides a number of desirable properties:

  1. it clearly represents the intended state of the infrastructure – no for loops, http calls, etc to interpret

  2. it aligns with how tools developed by the Kubernetes project are writtenkubectl, kustomize, etc

  3. it enables composition of different types of tools written in different languages

    • any modern language can manipulate YAML / JSON structures, no need to adopt go
  4. it supports static analysis and validation

    • develop tools and processes to perform validation and linting
  5. it supports programmatic modification

    • develop CLIs and UIs for working with configuration rather than using vim
  6. it supports customizing generated resources so the templates don’t need to be modified

    • artifacts generated from templates or DSLs may be modified directly, and then merged when they are regenerated to keep the modifications.
  7. it supports display in UI and tools which use either OpenAPI or the YAML/JSON directly.

Q: Isn’t writing YAML hard?

A: kpt offers a collection of utilities which enable working with configuration programmatically to simplify the experience. Using vi to edit YAML should be necessary only for bootstrapping, and the common cases should use setters or functions to generate or modify YAML configuration.

Q: I really like DSL / templating solution X. Can I use it with kpt?

A: Yes. kpt supports plugging in solutions which generate or manipulate configuration, e.g. from DSLs and templates. This may be performed using source functions from our catalog. The generated output may be modified directly, and merged when regenerated.

Q: I want to write high-level abstractions like CRDs, but on the client-side. Can I do this with kpt?

A: Yes. kpt's architecture facilitates developing programs which may generate or modify configuration. See the Functions User Guide for how to compose multiple programs together.

Q: How do I roll out changes throughout my organization using kpt?

A: This can be done one of several ways, including: 1) using semantic versioning or release channels with functions, or 2) updating packages.

Q: Is there a container image that contains kpt?

A: Yes. contains the kpt binary.

Q: How do I run source or sink functions?

A: All kpt functions, including source and sink functions, should be run using kpt fn run. The kpt fn source and kpt fn sink commands run builtin source and sink functions.

Q: I still have questions. How do I contact you?

A: Please reach out!

Next Steps

  • Learn about kpt architecture including major influences and a high-level comparison with kustomize.
  • Consult the command reference for how to use kpt.
  • Read kpt guides for how to produce and consume packages and integrate with a wider ecosystem of tools.