As far as Iâm aware there isnât a public API - not like Wanikani. I donât know which technologies are used, but from what I can tell the request to load the vocab pages returns the vocab embedded in page as the response - it doesnât seem like there is an endpoint for just getting the vocab. I could be wrong though?
Theoretically, you could try to parse out the vocab from these pages - if you could extract the vocab text, id, and deck from the page then you could get the list of vocab. Then it would be possible to hit the mark as known endpoint with the deck/vocab ids - the API isnât really designed for this though.