Static typing support

[multimeric]

I think it would be great if the full Seqera API were converted to Python types (dataclasses, TypedDict etc). This would not only be more user friendly than executing the tw CLI via strings, but would also catch usage errors statically.

I envisage being able to do something like:

from seqerakit.models import Workspace, Client

client = Client()
workspace = Workspace(name="showcase", organization="seqerakit_automation")
client.create(workspace)

In theory this could be done by generating code using openapi-python-client as recommended in the Seqera API docs, but this doesn't fully work: https://feedback.seqera.io/feature-requests/p/fix-openapi-schema.

[andrew-dawson-seqera]

Thanks for the feedback! I understand the value of having Python-native support for interacting with the Seqera API. While our current tool heavily relies on the tw CLI, we recognize the advantages of a more robust Python SDK.

At this time, our plans for the future of this tool are still being discussed, and we’re exploring various options, including the development of a proper Python SDK that would provide a more user-friendly interface similar to what you are describing.