cloudcoil-models-kpack

Versioned kpack models for cloudcoil.

🔧 Installation

Using uv (recommended):

# Install with kpack support
uv add cloudcoil.models.kpack

Using pip:

pip install cloudcoil.models.kpack

Usage

Cloudcoil 0.7 generates a typed lookup function so callers do not need to depend on schema-derived module names:

from cloudcoil.models.kpack import get_model

Image = get_model("Image", api_version="kpack.io/v1alpha2")
resource = Image.model_validate({
    "metadata": {"name": "example"},
    "spec": {'tag': 'example.com/app:latest', 'builder': {'name': 'builder', 'kind': 'ClusterBuilder'}, 'source': {'git': {'url': 'https://example.com/repo', 'revision': 'main'}}},
})
resource.create()

Generated resources support validation, fluent builders, and the Cloudcoil client API.

Development

uv sync --dev
make gen-models
make lint test
uv build

Generation uses the namespace and input configuration in pyproject.toml, with automatic resource identity and field alias inference. Generated modules are built on release branches; pull requests regenerate and test them before publishing.

schemas/lifecycle.json supplies the lifecycle API definitions omitted from the upstream v0.18.0 OpenAPI document, matching the upstream Go types. Other Kubernetes types are generated with this package, so a separate Kubernetes model dependency is unnecessary.