Weekly Review Rules

A task system without a weekly review degrades no matter how good the daily habits are — stray captures pile up, priorities go stale, and slipped items quietly vanish. The review is the one recurring step that resets the whole system to trustworthy, and it only works if it's fast enough to actually happen every week.

Fixed slot

• Put the review on the calendar at a specific recurring time (same day, same hour) rather than "whenever there's time" — a review with no fixed slot competes with everything else for attention and reliably loses.
• Protect the slot the same way a real meeting would be protected — treat a skipped review as a missed commitment, not a free pass, because a system that reviews every other week degrades noticeably faster than one reviewed weekly.
• Pick a time with low interruption risk (end of week, before other people's Monday-morning chaos starts) — a review interrupted halfway through rarely gets finished later the same day.

Empty all inboxes

• Process every inbox to zero before doing anything else in the review: email, task capture inbox, notes app, physical inbox, any place stray items land during the week. An unprocessed inbox at review time is exactly what the review exists to clear.
• "Empty" means every item has been triaged to a decision (do it, schedule it, delegate it, delete it, or file it as reference) — not archived unread, which just moves the pile somewhere less visible.
• Do this step first, every time, before reviewing what's already on the task list — reviewing an organized list while an inbox still has undecided items sitting behind it gives a false sense of being caught up.

Review the past week honestly

• Go through what was actually done versus what was planned, and note both categories explicitly — reviewing only what got done skips the more useful signal, which is what didn't and why.
• For everything that slipped, name a real reason (underestimated the time, got deprioritized by something more urgent, was never actually going to fit) rather than a vague "didn't get to it" — the reason is what determines whether it should be rescheduled, rescoped, or dropped.
• Notice patterns across weeks, not just this week in isolation — the same type of task slipping three weeks running means the estimate or the priority is wrong, not that willpower is.
• Don't skip this step when the week went well — a week that went to plan is worth reviewing too, because it tells you which planning assumptions actually held.

Plan next week's three outcomes

• Pick a small number of specific outcomes for the coming week (three is a workable default) — not a task dump, but the handful of things that would make next week a genuine win if nothing else got done.
• State each outcome as a concrete result, not a vague direction: "Ship the v2 draft to the client" not "make progress on the client project." A vague outcome can't be checked off honestly at the next review.
• Derive the coming week's priority tasks from these outcomes, not the other way around — decide what matters first, then let the task list serve it, rather than picking priorities out of whatever's already sitting on the list.

Calendar-vs-tasks reconciliation

• Look at next week's calendar and the task list side by side — a task list that ignores three days of back-to-back meetings is planning for a week that doesn't exist.
• Flag any task that has no realistic time slot given the calendar as it stands, and either move it, delegate it, or consciously accept it won't happen — an unreconciled conflict just becomes next week's unexplained slippage.
• Check the calendar itself for missing prep: a meeting with no agenda drafted, a trip with no packing task created. The calendar generates tasks the task list doesn't yet know about; reconciliation surfaces them now instead of the morning of.

Keep it under an hour

• Timebox the whole review, and if a section is running long (usually inbox processing or a rabbit-hole on one project), cut it off and finish the remaining sections at their normal pace — a review that regularly runs over gets skipped the next time something else is competing for the slot.
• Keep the review at the level of decisions and outcomes, not deep work — if a specific problem surfaces that needs real thinking, capture it as its own task for later rather than solving it inside the review.
• If the review is consistently taking too long, the problem is usually system debt built up during the week (an inbox that wasn't triaged daily, tasks that were never next-action-ized) — fix that upstream habit rather than accepting a permanently longer review.