Feature Request: Auto upload local code for remote workers (via `--working-dir`)

[dragongu]

Describe the current behavior

In remote-worker mode, worker machines cannot access the deploy client's local path.
Users must manually package/upload code to shared storage first, then configure pull steps.

This creates operational friction, especially for teams migrating from Ray workflows (ray job submit) where local code submission is typically one step.

Describe the proposed behavior

When deploying to remote workers, support an opt-in flow based on --working-dir:

prefect deploy ./flows/my_flow.py:my_flow \
  --name my-flow \
  --pool remote-pool \
  --working-dir /opt/prefect/flows

Behavior:

• Prefect automatically packages/uploads local code to configured shared storage.
• Deployment automatically gets pull metadata and working directory.
• Worker pulls and runs directly.

Storage location/credentials should be configured in Prefect server and worker startup config, so end users do not need to manage upload details in each deployment command.

Example Use

No response

Additional context

No response

[desertaxle]

Thanks for opening an issue @dragongu! I think what you're looking for currently exists as push steps. Push steps are essentially the mirror of pull steps and allow you to define where code should be uploaded when deploying your flow. This requires configuring an object storage bucket in a service like S3, GCS, or Azure Blob Storage if you want to use one of the push steps available in prefect-aws, prefect-gcp, or prefect-azure, respectively. You can also write a custom push step if you want to use a different storage provider.

Let me know if push steps don't fully satisfy your use case, and we can dig into what enhancement is necessary, but I think Prefect already has what you need!

[dragongu]

Thanks for opening an issue @dragongu! I think what you're looking for currently exists as push steps. Push steps are essentially the mirror of pull steps and allow you to define where code should be uploaded when deploying your flow. This requires configuring an object storage bucket in a service like S3, GCS, or Azure Blob Storage if you want to use one of the push steps available in prefect-aws, prefect-gcp, or prefect-azure, respectively. You can also write a custom push step if you want to use a different storage provider.

Let me know if push steps don't fully satisfy your use case, and we can dig into what enhancement is necessary, but I think Prefect already has what you need!

@desertaxle Thanks for the reply — I really appreciate you taking the time.

I did look at push steps before opening this issue. My concern isn’t that there’s no way to get code onto workers, but that it still feels operationally heavy for our team: in practice it pushes a lot of “packaging / upload / wiring” work onto every developer who deploys a flow.

Most of our engineers are already used to Ray’s ray job submit + --working-dir model, where specifying a working directory is enough for the tooling to handle getting local code to remote workers in one straightforward step. I’m hoping Prefect can offer something similarly low-friction for remote workers — ideally with storage/credentials and defaults owned by the server/platform, so individual deploys don’t require each developer to manually manage the upload path unless they want the extra control.