Weekly Review & Battle Plan
A guided weekly review that turns last week's chaos into next week's three priorities.
Good plans are just good questions in the right order. These planning prompts turn vague goals into structured roadmaps — weekly plans, project kickoffs, code refactor strategies, and lesson reviews — with the follow-up questions built in.
A guided weekly review that turns last week's chaos into next week's three priorities.
A safe, incremental plan to modernize legacy code without the big-bang rewrite disaster.
A complete lesson plan from one topic: objectives, timing, activities, and three differentiation tiers.
A good plan is mostly good sequencing: goals before tasks, risks before timelines, decisions before meetings. The planning prompts in this hub encode that sequencing. The weekly review closes last week before opening next one; the refactor plan maps the code before touching it; the decision matrix forces criteria into the open before options get scored. To get the most from them, answer the prompt's clarifying questions honestly instead of skipping ahead — the model's plan can only be as real as the constraints you admit to. And revisit: a plan pasted back into the same chat a week later, with what actually happened, becomes a review that makes the next plan smarter.
A plan you generate once and admire is decoration. The planning prompts in this hub are built to be re-run, and the re-running is where their value hides. Settle on a cadence and keep it boring: goals get set once a quarter, milestones once a month, and the weekly review — the engine of the whole system — every Friday or Sunday, same time, same chat. Paste in last week's plan plus a brutally short account of what actually happened. Three lines is enough: finished, slipped, surprised by. The model then plans next week against reality instead of against the fantasy version of you it met in January.
That plan-versus-reality gap is the most valuable input you will ever give a planning prompt. "I said I'd write four mornings; I wrote one, because client calls ate the rest" produces a next plan with protected mornings and a smaller promise. Skip the confession and you'll get the same optimistic four-morning plan again, and again, each time feeling like fresh motivation and each time failing the same way. Honest inputs are what separate a planning system from a horoscope.
Keep the layers separate in the chat. Goals live at the top and change rarely; when a weekly plan starts arguing with a quarterly goal, that argument is information — either the week is drifting or the goal was never real, and the prompt's clarifying questions will surface which. Daily work comes last and stays small: pull exactly three items from the weekly plan each morning, no more, and let the plan absorb the guilt of everything not chosen. People abandon planning systems because every day feels like failure against a long list. A weekly plan with daily top-threes turns the same work into a streak of kept promises — which is what keeps the system, and you, running past week three.
Build the bad week into the plan before life does it for you. Plan at roughly seventy percent of your real capacity — the remaining thirty is not laziness buffer, it's where the dentist appointment, the broken build, and the client emergency are already scheduled, whether you wrote them down or not. Then define, in advance, the minimum viable version of the week: the two outcomes that would still make it a success if everything else collapsed. When the bad week arrives — it arrives — you execute the fallback you chose calmly on Sunday instead of improvising cuts under stress on Wednesday. People experience the two situations completely differently. One feels like a professional adjusting; the other feels like failing, and the felt difference is what decides whether the system survives to week four.
Protect the ritual itself the way you'd protect a meeting with your most important client, because that's the honest description of it. Twenty unbroken minutes, phone in another room, same document open. A review done in fragments between messages produces a plan written by your interruptions. The week deserves better authorship than that.
More honest than detailed. Deadlines, real constraints, and what failed last time matter more than polished wording — the prompt structures the mess; you supply the truth.
It can draft the full structure — phases, milestones, risks, dependencies — in one pass. Treat it as a strong first draft: you validate estimates and ownership, it holds the shape. Revisit weekly with progress notes.
Claude handles long, document-heavy planning well; ChatGPT is excellent at structured tables and checklists. The prompts here work on both — consistency of use beats tool choice.
Weekly for active projects — the review cadence matters more than the plan's detail. Monthly for goals, quarterly for the big picture. A plan reviewed four times beats a perfect plan reviewed once.