Plan Your Week in One Prompt: The Weekly Planning Formula

Productivity Updated August 11, 2026 6 min read

Four blank white index cards laid in a row on a pale wooden desk, a fountain pen resting below them and a green mug further back

What’s in this guide

The short version: a weekly planning prompt has five parts — the dump, the constraints, the sorting rule, the choosing rule, and the stop instruction. Leave any one out and you get a specific, predictable failure. This page is each part and what it prevents.

If you just want the finished prompt and the ten minutes around it, the Sunday reset is that. This page is for when the reply keeps coming back wrong and you want to know which line to change. The wider system both sit inside is the productivity guide.

Five parts, and the failure each one exists to prevent. The stop instruction is the one almost nobody writes, and it is why most replies arrive as a schedule nobody asked for.

Part one — the dump goes in raw

Paste your list exactly as you wrote it. Misspellings, half-sentences, three things on one line. And tell it not to tidy: I will paste a messy list; do not tidy it up.

What it prevents: a reply that reformats your list and hands it back. Without that line you reliably get your own forty items returned in tidy sentence case, which feels like progress and has changed nothing. You did not need a typist.

Part two — the constraints, in numbers

Three lines: how many days are already committed and to what, roughly how many hours you actually control, and what will interrupt you.

What it prevents: a plan for somebody with an empty calendar. This is the part people skip because it feels like admin, and it is the part that decides whether the answer is about your week or about a generic week. Hours I actually control is the highest-value line in the whole prompt.

Part three — the sorting rule

Ask for three named buckets — must happen this week, would be good, someday — and add the clause that does the real work: move anything vague into someday and say why it went there.

What it prevents: the permanent residents. Every list has four or five items that have been carried for months — “sort out the insurance”, “look into the thing” — which are not tasks, they are unmade decisions wearing task clothing. Naming them as vague and explaining why is often the most useful sentence in the reply.

Part four — the choosing rule

Ask for three, ask for one line of reasoning each, and ask for the drop order.

What it prevents: a ranked list of eleven things, which is the same as no list. Asking for reasoning matters more than it sounds — it makes the choice inspectable, and about a third of the time the reasoning is obviously wrong in a way the ranking alone would have hidden.

The drop order is the part that survives the week. Ask which of the three goes first when things go wrong and why that one.

Part five — the stop instruction

Two lines: do not give me a schedule, I will place them myself, and if the must-happen list clearly does not fit the hours I said I have, say so plainly rather than making it fit.

What it prevents: two things, and they are the two failures that make people give up on planning this way.

The first is the hour-by-hour timetable. Ask a model to plan a week and it will happily produce Monday 9:00–10:30 with no knowledge of when your meetings are, and a timetable that is wrong in three places is harder to fix than a blank page.

The second is worse because it is invisible. Given more work than hours, the default behaviour is to fit it in — smaller estimates, tighter gaps, a plan that looks reasonable and is not. Asking it to say so plainly converts a silent impossibility into a sentence you can act on.

The whole thing

The weekly planning formula

Help me plan the coming week. I will paste a messy list; do not tidy it up. MY LIST: [paste everything on your mind, unsorted] Days already committed: [how many, and to what] Hours I actually control: [a rough number, minus a third] Interruptions I should assume: [children, on-call, commute, whatever is true] Do this in order: 1. Group my list into MUST HAPPEN this week / WOULD BE GOOD / SOMEDAY. Move anything vague into someday and say why it went there. 2. From MUST HAPPEN, name the THREE that would make the week feel handled. One line each on why. 3. For each of the three, say roughly how long it needs, and whether it can be done in interrupted chunks or needs a clear run. 4. Tell me which one to drop first if the week goes wrong, and why that one. Do not give me a schedule. I will place them myself. Be honest: if the must-happen list clearly does not fit the hours I said I have, say so plainly rather than making it fit.

Adapting it to something that is not a week

The five-part shape holds for any planning job. Change the horizon and the buckets and the rest stays.

A single day

Buckets become before lunch / after lunch / not today. Choose one thing rather than three. The drop rule matters more, not less, because a day has no slack in it at all.

A quarter

Buckets become this month / next month / not this quarter, and “hours I control” becomes weeks. At that horizon the useful output is decomposition rather than selection — breaking one goal into twelve weeks is the version built for that.

Somebody else’s week

It works, and be careful with it. A plan made for another adult without them in the room is a plan they did not agree to. It is genuinely useful for a shared week — the household one — provided the choosing happens with both people present.

What this will not do

It does not know what matters to you and it will sound confident regardless. Every judgement about importance in the reply is a guess dressed as an answer, which is exactly why part four asks for reasoning — the reasoning is checkable and the ranking is not.

What it is genuinely good at: grouping forty unordered items in seconds, noticing the ones that have been vague for months, spotting that three of your items are the same task, and telling you the arithmetic does not work. That is a real contribution and it is not the same as knowing your priorities. If the list is mostly invisible household work, the mental load version is the better starting exercise.

The printable

The Weekly Planner has the five parts on the back as a card, so writing your own version does not mean looking this page up again.

Questions about the planning formula

Why does it keep giving me an hour-by-hour schedule?

Because that is the default shape of an answer to “plan my week”, and it has no idea when your meetings are. Add do not give me a schedule, I will place them myself as its own line near the end. If it still does, reply with just the three and the drop order please.

How accurate is the hours estimate?

Treat it as a rough ordering rather than a measurement — it does not know how fast you work. What it is reliable at is relative sizing, telling you that one of your three is much bigger than the other two, which is usually the thing you needed to see.

What if it says my week does not fit?

Believe it, and check the arithmetic yourself. That answer only appears because you asked for it explicitly, and the default alternative was a plan that quietly assumed you would find the hours. Drop something on Sunday rather than failing at it on Thursday.

Can I reuse the same prompt every week?

Yes, and that is the intended use. Save the filled-in version with your constraints already written and paste it each Sunday — the constraints are the part that took the time, and they rarely change week to week. Update the hours line when your circumstances do.

Is a prompt better than just thinking about it?

For choosing, no — you are the better judge and the reply is a guess. For sorting forty unordered items, spotting the vague ones, and doing the arithmetic on whether it fits, yes, and those three take most of the time when you do it in your head.