1. make sure you have node installed
2. run npm install
3. install BBT in Zotero. Right-click a collection or library and select "Download Better BibTeX export", and choose the "CSL JSON" format.
4. edit config.json. Paste the URL from step 3 into the field collection. Also paste the URL in your browser so you can see what the objects look like
5. edit template.html. The template understands nunjucks syntax; the objects it inserts are in CSL-JSON format, the fields of which are documented here. Variables with dashes (-) are available in camelCase in the template, so a variable like container-title is containerTitle in the template.
6. create a directory called output if it doesn't exist, or add a field called output to config.json pointing to the output directory
7. run npm start

You can see what data is available to the template by plugging the URL from step 3 into your browser. this script will make the following changes to what you see there:

1. relations is removed
2. collections is removed
3. if a DOI is present but starts with https://doi.org/, that prefix is stripped
4. links is the subset of attachments that are links to URL pages
5. attachments will have all links removed
6. a field named select will be added which is an URL that directly links into your Zotero library
7. creators will be changed into a single field listing all creators
8. tags will be changed into a single field listing all tags, sorted alphabetically
9. notes will be changed into a single field with all notes

Nunjucks will escape everything so that it is valid HTML; there is one special field called notes which is already valid HTML, so if you want this in the template, add it as {{ notes | safe }} so nunjucks doesn't escape it again