Weekly Planning with AI
AI is useful for the mechanical parts of weekly planning: structuring a brain dump into organised categories, writing status updates from bullet notes, and identifying conflicts in a proposed schedule. It cannot make prioritisation decisions — it does not know your manager's expectations, your team's current priorities, or what is actually urgent versus what feels urgent. Use it to speed up structure; make decisions yourself.
Friday Weekly Review Template
A consistent end-of-week review prevents the "what did I even do this week" problem and makes Monday planning faster. Build a template with AI once; reuse it every Friday.
Create a weekly review template for a [DESCRIBE YOUR ROLE — e.g., "senior software engineer in a product team"].
The review should take 15-20 minutes and produce:
1. What I completed this week (in a format I can paste into a status update)
2. What I planned but did not complete — and why (one line per item)
3. Blockers I need to surface to my manager or team
4. Wins worth acknowledging (even small ones — for mental health and future review)
5. What I learned this week (one or two things maximum)
6. Top 3 priorities for next week (not a full task list — just the three things that matter most)
Format it as a fill-in template with [BRACKETS] for my inputs. Keep each section brief — the review should not become an essay.
Run this template generation once. Save the output in a notes app and copy it every Friday. The value is in the habit, not the tool.
Monday Priority Planning
Monday planning with AI is most effective when you paste your actual task backlog and let AI organise it — not when you ask AI to plan your week from scratch without context.
Help me plan my week. I will provide my task list and context; you organise it.
This week context:
- Top priority from my manager: [1-2 SENTENCES]
- Commitments I cannot move: [MEETINGS, DEADLINES, ON-CALL]
- Hours available for focused work: [e.g., "approximately 20 hours — rest is meetings"]
- Energy pattern: [e.g., "best focus in the morning; afternoons suit meetings and reviews"]
My task list (paste everything — do not pre-filter):
[PASTE YOUR FULL BACKLOG]
Group tasks into:
1. Must-do this week (mission-critical, deadline, or manager-priority)
2. Should-do this week (important but moveable if needed)
3. Defer to next week (no consequence to waiting)
4. Delete or delegate (not actually your responsibility)
Do NOT assign specific days or times — I will do that. Just prioritise and group.
The explicit "do not assign times" instruction is important. AI time assignment ignores energy levels, meeting context, and task dependencies you know but have not written down.
Weekly Status Update
AI converts rough bullet notes into a clean status update much faster than writing from scratch. Paste your raw notes; provide the audience and format.
Convert these rough notes into a weekly status update. Audience: [MY MANAGER / MY TEAM / STAKEHOLDERS].
Format required: [e.g., "bullet points under three sections: Completed / In Progress / Blockers. Max 150 words total."]
My notes from this week:
[PASTE RAW NOTES — incomplete sentences, mix of tasks and context is fine]
Rules:
- Completed items: past tense, specific (mention feature or ticket name if present in notes)
- In progress: include expected completion if I mentioned a date
- Blockers: be specific about what is blocking and what I need from whom
- Do not add information I did not include in the notes
- Do not use jargon unless it appears in the notes
"Do not add information I did not include" prevents AI from filling in plausible-sounding but fabricated details about your work.
Workload Conflict Detection
I have the following tasks and commitments for this week. Identify conflicts and capacity problems.
Fixed commitments (meetings, on-call, deadlines):
[LIST WITH TIMES AND DURATIONS]
Tasks I have committed to completing this week:
[LIST TASKS WITH ROUGH ESTIMATES IF YOU HAVE THEM]
Available hours for focused work this week: [TOTAL HOURS]
Identify:
1. Are there more committed hours than available hours?
2. Which tasks have no buffer if they take longer than estimated?
3. What must I communicate to stakeholders if I am over capacity?
4. What should I ask to defer or descope?
This prompt surfaces the overcommitment conversation you need to have on Monday rather than Friday when it is too late to recover.
Common Planning Patterns
| Task | What to give AI | What to decide yourself |
|---|---|---|
| Priority grouping | Full task list + context about manager priority | Final order within each priority group |
| Status update | Raw notes + audience + format constraints | Review for accuracy before sending |
| OKR progress update | OKR text + this week's work that relates to it | Whether progress is on track (you know the real context) |
| Meeting agenda | Meeting goal + topics to cover + time available | Which topics actually need this group versus async |
| Quarter review preparation | Bullet list of accomplishments + your role's goals | Which accomplishments are most worth highlighting |
The Prioritisation Limit
AI cannot prioritise for you
Prioritisation requires knowing: what your manager cares about most right now, which dependencies will block teammates if you delay them, which tasks have invisible deadlines (stakeholder expectations not written down anywhere), and your own energy levels and cognitive load. None of this is in your task list. AI can organise your tasks by category; only you can rank them correctly. Treat AI priority groupings as a draft to revise, not a plan to execute.
Checklist: Do You Understand This?
- What is the primary value of a weekly review template — and why build it once rather than prompt AI each week?
- Why should you instruct AI not to assign tasks to specific days or times?
- What two rules prevent AI from fabricating information in your status update?
- What does a workload conflict detection prompt surface — and when is the right time to run it?
- Write a Monday priority planning prompt for your current role, including the context fields that AI needs.
- Why is prioritisation a task AI cannot perform correctly even with access to your full backlog?