Create an ingests.yaml, single python command to run all ingests

[kevinschaper]

We need our source of truth to list all of the ingests, and rather than enumerating all of the ingests in the Jenkins file - we need to wrap the process to execute them in a single python call.

In addition to the bare minimum of executing, it would be good if the ingests can run with some parallelism, and also as a stretch goal, if we can separate out the logs.

I'm going to take a shot at using Prefect for this, but I'll wrap it in a Typer CLI. I also plan to remove the Dagster pipeline code alongside this change.

[kevinschaper]

With respect to this issue, the output of the merge step should be:

• nodes tsv, all of them
• edges tsv, where the subjects & objects point to nodes we have
• edges tsv, where either a subject or object points to a missing node

We'll do stats in a different issue after this goes in

[kevinschaper]

Implemented as pipeline.py