Deciding what to eat, taken off the table

The problem

Deciding what a family eats this week is a small decision repeated until it becomes a chore. Most people solve it badly: the same six dinners on rotation, a shopping trip that forgets an ingredient, and food that goes off in the back of the fridge.

The meal-planning apps that exist mostly hand you a recipe database and leave the planning to you. That is the hard part, and it is the part nobody wants to do on a Sunday evening.

What I built

Madplanen generates a full week of dinners in one click. If a meal does not appeal, it is swapped for another with a single click rather than sending you back to a blank plan.

Every dish carries its protein, carbohydrate, fat and fibre, so daily targets are tracked as a by-product of eating rather than as a second app to maintain. Weekly weigh-ins are logged and charted, with a Sunday reminder to keep it honest.

The finished week collapses into one shopping list, grouped by supermarket category so it can be walked in aisle order. Recipes can be browsed, written or imported, each with ingredients, cooking time and nutrition.

There is a deliberate cheat-meal slot in every week. A plan people abandon after nine days is worth nothing, so the product builds in the exception rather than treating it as failure.

Where it stands

Madplanen is live at madplanen.net. I co-founded it with Lucas, and I write and maintain the code.

It runs on Blazor Server, which keeps the interactive planning UI in C# rather than splitting the logic across a separate front-end — a reasonable trade for a small team where every context switch costs real time.