Grida

Grida is an early-stage, node-based workflow orchestration SDK for digital content creation (DCC) pipelines.

Version 0.0.1 provides a small, working graph kernel for declaring dependent operations, validating the graph, and executing nodes in deterministic topological order. It has no runtime dependencies.

Early preview: This release establishes the package name and the first usable API. The API may change substantially before 1.0.

Installation

python -m pip install grida

Example

from grida import Node, Workflow

workflow = Workflow("character-export")
workflow.add(Node("scene", lambda inputs: "hero.ma"))
workflow.add(
    Node(
        "export",
        lambda inputs: inputs["scene"].replace(".ma", ".fbx"),
        needs=("scene",),
    )
)

outputs = workflow.run()
assert outputs["export"] == "hero.fbx"

A node receives a read-only mapping containing the outputs of its direct dependencies. In a real DCC integration, its operation can call the host API, launch a subprocess, or invoke an adapter captured by the operation.

Current scope

Included in 0.0.1:

• Node and workflow declarations
• Dependency validation
• Cycle detection
• Deterministic in-process execution
• Typed exceptions for invalid graphs

Not yet included:

• Maya, Blender, Houdini, Unreal, or other host adapters
• Persistent graphs or a graph file format
• Distributed execution, caching, retries, or a scheduler
• Stability guarantees for the public API

Development

Run the tests with the Python standard library:

python -m unittest discover -s tests -v

License

MIT