This is interesting. So it's HTML -> PDF without using Chrome/Puppeteer etc.? Is that correct. It reminds me of
https://weasyprint.org/Weasyprint (by their own admission) have really struggled with a lot of bugs, which isn't surprising if you're allowing arbitrary HTML input - how are you getting on with that?