I mean the llm you used is a saas. But the question is whether a framework for doing this has value (I think it does)

▲ wewewedxfgdf 4 days ago | parent [-]

I then added on an SMS message that tells me things I COULD eat at every meal time. It does not say what I must eat, only presents options - I don't like being told what I must eat, all I want to know is what's the best choices of what I have, then I can choose.

  Time for Lunch.
  Choose one of:
  ---
  Tuna salad: 1/2 cup tuna (drained) + 3 cups salad greens + 1/2 cup beetroot + vinegar/lemon (≈850 kJ)
  ---
  Egg salad: 2 boiled eggs (chopped) + 3 cups salad veg + 1/4 cup grated cheese (≈1000 kJ)
  ---
  Chicken salad bowl: 1/2 cup cooked chicken pieces + 3 cups salad veg (≈950 kJ)
  ---
  Minestrone soup: 2 cups (≈1060 kJ)
  ---
  Creamy chicken soup: 2 cups (≈1090 kJ)

	▲	mips_avatar 3 days ago \| parent [-]
		I think it's kind of like AI coding, you can do an awful lot just in the chat with chatgpt but there's a lot of advantage of a framework like cursor/claude code to manage more complicated context and run things for you. For this case maybe recipe gen is something you want more control over and are ok managing the context of things in your pantry and other stuff in your head. But I think if the context of things like your pantry was indexed really well there'd be a huge advantage in having a framework beyond chatgpt messages.