develop github workflow for assessment process

[chantelwetzel-noaa]

I mentioned this on our most recent meeting and the yellowtail assessment team recently started a similar discussion that I thought included useful information.

• we should have separate R scripts for each data type or data source (e.g. commercial comps, survey comps).
• each script should result in a dataframe (or dataframes), which are written either a .rda file that is loaded as a data object in the package or saved as a csv in the data-processed folder that is written and formatted for SS3.
• a single script can read each of the model data files that can be used to modify SS3 models.

The repository for the 2023 update assessment for sablefish provides a useful example that we can modify as we see fit.