I am trying to use deployments to create generic pipelines that can be used by different projects, but I can't find a simple way to customise the infrastructure, e.g. volume mounts for containerised deployments to give data pipelines access to the project-specific input and output directories. A way to do this at deployment runtime would be nice. I'm also open to creating project-specific deployments of generic flows, but the projects don't have the pipeline dependencies, so can't deploy the pipelines, and I can't see how to do this through the python SDK, e.g. read generic pipeline, edit infra, create new deployment. Another idea was to start workers with custom base job templates which define the volume mounts using

prefect worker start --base-job-template path/to/bjt.json

, however this doesn't seem to affect the runs at all, they don't get the volume mounts.