First, what is the benefit of using this over something established like fddb or any already existing nutritional app?

Second, just translating the words for the units (like cup) doesn't work, because for example while "szklanka", the Polish word for glass, is indeed used in Polish recipes, there it means more something along the lines of "just grab whatever standard glass you have, it will be fine". It explicitly does not mean a unit of measurement of 236 ml.

Third, since you are using a US source you fall into the trap of US defaultism here. The way carbohydrates are labeled is different around the world and I think only the US does this weird thing where you have to manually subtract the dietary fiber from the total carbohydrates. This matters because you essentially give non-US users (and definitely users from the EU) plainly incorrect amounts of labeled carbohydrates.

All in all, cute project and good on you trying to tackle it. It looks okay though the AI smells bother me personally, but I don't think nutritional data is a great place to start for such a project and honestly I will not remember your site in a few days anymore.

In such a framing I don't find it surprising at all that teachers prefer the more polished answers generated by AI, because if LLMs are good at one thing, it is being confident in whatever they generate and present it convincingly.

Is my understanding correct that a normal LLM is stateless in the sense that when you talk to it today about frying pans, it does not remember that you spoke about fried rice yesterday? Is this solution effectively adding Markdown files as part of the prompt? Essentially writing into a file "whenever I talk about scripting, I explicitly mean the zsh"?

Relevant XKCD: https://xkcd.com/1053/

Who?