Clarify how inputFileUrls are handled for engines other than Cromwell

[nh13]

See specific instructions for Cromwell here: https://aws.github.io/amazon-genomics-cli/docs/concepts/workflows/#manifestjson-structure

How is this handled for miniWDL and others? I can see wanting having two input JSON files, one for a static set (e.g. reference genome files) across workflows, and one for sample-specific (changes across invocations of the workflow), specifically for miniWDL. Are these JSONs merged prior to handing off to miniWDL for example?

[patmagee]

it would be great to have more clarity here as well. @nh13 I had to go through the agc codebase to determine how non cromwell backends handle workflows. It turns out if you want to use the WES API directly, you need to upload the workflow.zip to S3 directly and use the s3 object url as the workflow_url.

It would be nice if the WES api for miniwdl (and other backends) was able to handle the exact same request