The Docs section currently repeats full text of each citation in each markdown document. This can be handled better with a data collection (YAML or JSON) and by referencing the an id for each citation. For instance:
You can also export BibTex to JSON with Pandoc
pandoc filename.bib -t csljson -o data/filename.json
Store data according to CSL-JSON spec and then render that in Hugo with Shortcodes and Partials