Hello, we are looking to use the S3 storage/ecs run config combination. We have a process where we register a flow to a UAT project on merge to master and register the flow to a production project on release. We want to leverage the stored as script functionality. What I want to do here is on the merge to master push the flow .py file to S3 manually and then register the flow without having the storage object push it while still having the flow run use this artifact I pushed. This is helpful so that when I release...the artifact for the commit id I am releasing is already in S3 and I can then register the production flow using that same flow .py file that is already in S3. I was trying to make sense of

If neither are set then script will not be uploaded and users should manually place the script file in the desired key location in an S3 bucket.

I could not gather how to make sure the flow run will use the S3 artifact.

This is correct (we should update the docstring to be clearer here). If you set

stored_as_script

but not

local_script_path

, then prefect assume's you've put the file there yourself. Only if

local_script_path

is set will prefect upload the file for you (this is also automatically done if you use

prefect register flow

).