We need some kind of automated superhighway to quickly iterate on TAP_SCHEMA changes, and keep these in date with the felis files.  In short, we want to automate this pipeline:

1. PR to the https://github.com/lsst/sdm_schemas with a change to a felis yaml file.
2. Once merged, we allow GPDF to make a github release with a version number, that creates a tag and runs a publishing pipeline github action.
3. This will run the python felis library against the yaml files in the sdm_schemas directory and make different docker images for the different environments
4. Push the images to dockerhub
5. Make a commit for the TAP_SCHEMA chart to update the version and release a new chart
6. Argo-cd will have a new app that just has TAP_SCHEMA database, and break that off from the normal TAP service, since it's the data part that changes frequently
7. Argo-cd will be set to autosync only the TAP_SCHEMA app so we can test out how that goes.  For some environments we can also have manual intervention.