Imports

In SPX SDK docs, “imports” covers the mechanisms used to bring Python code into a running SPX configuration.

Use imports when YAML alone is not enough:

• loading custom Python classes from local .py files (quick prototyping, tests),

• binding Python object properties/methods to SPX attributes,

• or wiring lifecycle methods (start/run/stop) into a model definition.

Two common patterns

1. Registry-based components (recommended for libraries)

Package your extension code as a Python module and register classes via the SDK registry. This is the most maintainable path for reusable actions/components.

See: Registry.

1. File-based imports with python_file / import

Use the python_file component (alias: import) to load classes from repo-local files and bind them into the runtime.

See: PythonFile.

Minimal example (YAML)

python_file:
  extensions/my_block.py:
    class: MyBlock
    attributes:
      temperature_c:
        property: temperature_c
    methods:
      run:
        method: tick
        args: [1]

Notes:

• Prefer repo-relative paths checked into your project for reproducible builds.

• The path must exist inside the SPX Server process (in Docker, mount the folder into the container).

• args/kwargs in methods can reference runtime values via attribute refs like $attr(value).