The main thing is to NOT load paged content, but EVERYTHING in your template (range site.Pages) and then parse it through whatever creates your content. Add title and content and maybe some data fields (dates, keywords). Go from there.
Now, for your case, what you are not handing over is the format you need. txt, pdf, md can be anything. Can it have links? then do an markdown export with proper formatting. Can it have text only? Text format obviously. PDF is overkill, but I assume (black-box for me) they don’t care about formatting and only want text and links.
Edit: my sample code creates a JSON file with all URLs in the resulting website. For testing. It’s minimal.