Better process for Static Finding Aids generation

[bibliotechy]

Static Finding Aids are still deeply imperfect. even for not huge ones.

Dev team discussion led to to idea for a process like

• Create a Airflow Dag for generating static finding aid HTML
  • Dag should spin up an arclight worker that has a full rails server in it
  • It makes a request to localhost for that static page
  • it saves the output of theat request to s3
  • In the normal guides pages, we link to the HTML in s3.