The actual problem
I had a 500+ recipe library in AnyList. The recipes were there. The problem was the weekly gap between "we should eat at home more" and actually knowing what to buy at Kroger on Saturday.
Every approach I tried had the same failure mode: the planning part happened in one place, the recipes in another, and the shopping list in a third. There was always a translation step where I had to hold everything in my head at once. That's the step that breaks down when you're tired or distracted or both.
I needed the AI to live where the system already lives, not pull me into a new one.
How it works
I built a Claude MCP integration that connects directly to my AnyList recipe library. On Saturday or Sunday morning, I tell Claude what sounds good this week, what we had recently, and what's already in the fridge. It pulls from the actual library, not a generic list, and proposes a week of dinners. I can push back, swap things out, say "something faster on Tuesday," and it adjusts.
When the plan is finalized, it generates a shopping list organized by section: produce, proteins, pantry, dairy. I add it straight to AnyList. Kroger pickup, Saturday morning, done before coffee's cold.
The AI isn't replacing a tool I love. It's adding a layer of intelligence on top of one I already use. The system meets me where I already am.
What changed
I stopped dreading Sundays. That's the unglamorous, completely real outcome.
More specifically: we cook at home more often because the friction of planning is gone. The mental load of "figuring out dinner" no longer leaks into the rest of the week. The shopping list is accurate because it comes from actual recipes, not my best guess at what I'll need.
How it was built
The hardest part was writing prompts that understood our actual preferences. Not dietary restrictions and cuisine types, but the real things: fast enough for a Tuesday, fancy enough to feel like Saturday, nothing with more than 40 minutes of active time.