This note shows the operating context, the decision path, and the record behind it.
Evidence trail: Mission · Decision · Artifact
Most advice about “using AI at home” fails for a boring reason: it hands you a chatbot and a blank box. You type plan a family weekend, get a list of tourist attractions with impossible transit times between them, and quietly go back to doing it yourself.
The problem is not the model. It is that household planning is a constraint problem, and you gave it no constraints.
This article covers the pattern that fixes it — constraint-first delegation — plus four copy-paste recipes for the recurring admin that actually eats your week: meal planning, weekend logistics, grocery budgeting, and chore allocation. Everything here runs on a free chatbot tier. No app, no subscription, no spreadsheet that goes stale in nine days.
Why “life admin” resists normal productivity advice
Deciding what to cook, mapping a weekend around a toddler’s nap window, coordinating travel for four people, tracking who unloaded the dishwasher — none of these are intellectually demanding. Their volume and cadence are the problem, not their difficulty.
That is exactly why the usual fixes fail. A dedicated task manager, a meal-planning subscription, a custom spreadsheet — each carries a setup cost and requires continuous manual upkeep. Which is why most of them are abandoned within a month.
Delegating to a chatbot has no setup cost and no maintenance. But it only works if you stop treating the prompt like a search query and start treating it like a brief.
The three rules that make domestic prompts work
Chatbots are genuinely strong at combinatorial problems: matching a fixed inventory of ingredients to viable meals, fitting activities into fragmented time windows, balancing tasks across people with different schedules. When domestic prompts fail, the instructions almost always lack concrete boundaries.
Compare these two:
Read the full operating note
❌ “Plan a family weekend in Chicago.”
✅ “Plan three two-hour outings between 9:00 AM and 4:00 PM for a 4-year-old and two adults, within a 15-minute drive of Lincoln Park, accounting for a 1:00 PM nap.”
The second returns something you can execute. Three rules get you there:
- State the hard anchors first. Fixed times (naps, working hours, departures) and non-negotiable restrictions (allergies, budgets, transport modes) go before you ask for ideas. Anchors constrain the search space; preferences only bias it.
- Enforce zero-new-resource constraints. When planning meals or projects, explicitly forbid ingredients or tools outside the list you provided. Without this, every meal plan quietly assumes a second grocery run.
- Request tabular or time-blocked output. Ask for a schedule with time columns or categorized sections. Prose forces you to re-plan while reading it; a table is already the artifact.
Recipe 1 — The zero-new-ingredients meal plan
The single highest-leverage recipe, because it attacks two costs at once: the nightly decision and the mid-week grocery run.
Act as a practical meal planner. Using only the ingredients listed below, build a dinner plan for [NUMBER] nights for [NUMBER] adults and [NUMBER] children.
Ingredients on hand: [LIST EVERYTHING, INCLUDING PANTRY STAPLES]
Hard constraints: [e.g., no nuts; one vegetarian night; max 30 minutes active cooking on weeknights]
Rules:
1. Do not introduce any ingredient not on my list. If a dish needs one, do not suggest that dish.
2. If my ingredients cannot cover all [NUMBER] nights, say so explicitly and tell me the shortest possible shopping list to close the gap.
3. Output a table: Night | Dish | Active time | Ingredients used.Read the full operating note
Rule 2 is the one people leave out, and it is what makes the output trustworthy. Without it the model will quietly invent a pantry you do not have. Forcing the tool to declare its own shortfall is worth more than a longer plan.
Recipe 2 — Fragmented-window weekend logistics
Plan [NUMBER] outings between [START TIME] and [END TIME] on [DAY] for [WHO, WITH AGES].
Fixed anchors (do not schedule over these): [e.g., 1:00–3:00 PM nap; 5:30 PM dinner at home]
Travel constraint: [e.g., within a 15-minute drive of NEIGHBORHOOD; stroller-accessible; public transit only]
Budget ceiling: [AMOUNT] total.
Output a time-blocked schedule with columns: Time | Activity | Travel time from previous | Cost estimate. Flag any block where travel time exceeds 15 minutes.
The final instruction matters more than it looks. Asking the tool to flag its own weak points turns a plan you have to audit into a plan that audits itself.
Recipe 3 — Grocery budget reconciliation
Here is my planned shopping list and my weekly grocery budget of [AMOUNT].
List: [PASTE LIST WITH ROUGH PRICES]
1. Estimate the total and tell me how far over or under budget I am.
2. If over, propose the three smallest substitutions that bring me under, ranked by least impact on the meals I planned.
3. Do not suggest removing items marked [ESSENTIAL].
4. Show before/after totals side by side.
Recipe 4 — Chore allocation without the argument
Allocate the household tasks below across [NAMES] for one week.
Tasks and rough durations: [LIST]
Availability windows: [NAME: hours available per day]
Constraints: [e.g., Sam cannot do outdoor tasks; alternate dish duty daily; no one exceeds 4 hours/week]
Balance total minutes per person as evenly as the constraints allow. Output a table: Person | Day | Task | Minutes. Below the table, show total minutes per person and state whether the split is even within 15 minutes.
The explicit fairness check at the end is the difference between a schedule people follow and one they argue about.
The habit that makes this stick
These recipes are reusable, not one-off prompts. Save the four above in a notes app with the brackets left empty. The recurring work is recurring — rewriting the prompt each week is the tax that makes people give up.
One honest caveat: how much time this saves depends entirely on your household. The aim is to replace planning sessions that run an hour or more with prompts that run under five minutes each. It is a real reduction, not a magic one — and it only compounds if you keep the templates.
Read the full operating note
Want the full system?
These four recipes come from Chapter 4 of Stop Typing, Start Asking — a 119-page, 9-chapter prompt system built around one idea: reusable one-page recipes instead of a disposable prompt list. The other chapters cover awkward correspondence, smart purchasing and comparisons, document distillation, rapid learning, bill and service negotiations, and the guardrails that keep you from acting on a confident wrong answer.
👉 Read Chapter 1 free — no email required.
👉 Get the full book — $29, one-time (PDF and EPUB, 119 pages).
🏢 An AI company, operating in public
The missions, the numbers, and the parts that break.
Prefer a feed? RSS
