What to hand over
- Turning a pile of tasks into groups and an order
- Estimating how long things take, which you then correct
- Drafting a first layout of the week
- Re-planning after a disruption
What to keep
- What matters. No AI knows your priorities better than you.
- What to drop. The best plans delete tasks.
- How full the week is. Left alone, AI will happily fill every hour.
The 20-minute Sunday routine
- Brain-dump everything on your mind (5 minutes).
- Pick your three outcomes for the week; everything else is secondary (3 minutes).
- Ask AI to draft blocks for those outcomes first, then fill smaller tasks (5 minutes).
- Move blocks by hand until they match your real week (5 minutes).
- Leave Friday afternoon open for overflow (2 minutes).
A prompt for the weekly draft
My three outcomes this week are: [1, 2, 3]. Fixed commitments: [list]. Spread focused work across Monday to Friday, max five focused hours a day, and leave Friday afternoon open. List anything you couldn't fit.
The midweek check
Plans meet reality by Wednesday. Spend two minutes seeing what slipped, then move it or delete it. Deleting is allowed; a lot of weekly planning is admitting what won't happen.
A sample week shape
| Day | Focus |
|---|---|
| Monday | Outcome 1 (deep work), plan the week |
| Tuesday | Outcome 2, meetings |
| Wednesday | Outcome 1 continued, midweek check |
| Thursday | Outcome 3, admin batch |
| Friday | Wrap up, review, open afternoon |
Weekly review questions
- What did I plan that I didn't do, and why?
- What took longer than expected?
- What gave me energy, and what drained me?
- What should I stop doing?
- What's the one outcome that would make next week a good one?
Theme days
Giving each weekday a rough theme (deep work, meetings, admin, planning) reduces switching costs. Ask the AI to respect your themes when it drafts blocks, and keep exceptions short.
A meeting-heavy week
- Put the meetings in first.
- Find the two or three longest gaps and protect them for deep work.
- Batch email into one or two slots a day.
- Cancel or shorten any meeting you can.
Common failure modes
- Planning five days at 100 percent capacity.
- Never looking at the plan after Monday.
- Treating the plan as a test you can fail instead of a tool you can edit.
- Adding tasks mid-week without removing any.
Planning this on a real timeline is easier than reading about it.
Try the free planner →Prompts for the weekly review
Here's what I planned and what I actually did this week: [paste]. Tell me where my estimates were furthest off, what I tend to skip, and one change that would make next week more realistic.
Feeding back real outcomes turns the AI from a generic planner into one tuned to how you actually work. It's the single habit that most improves the quality of later drafts.
Setting a weekly capacity
Decide in advance how many focused hours you can truly deliver in a week; for many people it is far less than the hours they're at work. Tell the AI that number and ask it to stay under it. The honest cap is what keeps a plan from being a wish list.
| Situation | Realistic focused hours per day |
|---|---|
| Many meetings | 1 to 2 |
| Balanced week | 3 to 4 |
| Protected deep-work days | 4 to 5 |
These are rough planning figures, not rules. Use your own numbers once you've tracked a couple of weeks.
Doing it in Fuidzy
Plan the week on a timeline you can drag, with habits already on the schedule. Ask @fui in plain English to move or add blocks. It asks first, so nothing moves without your say. Free to start at app.fuidzy.com.