How to plan an AI itinerary around opening hours

Tell AI about closed days, last entry, and seasonal hours so museum Mondays and early kitchens stop wrecking the afternoon.

A yellow and white columbine flower with long spurs against a black background

You walked up at 10:05 on a Monday with the museum already in the day plan, tickets screenshotted, and a lunch reservation waiting two neighborhoods away. The doors were locked. A small paper sign said closed Mondays. The model had written "morning at the museum" like every morning was the same morning. You killed forty minutes on a plaza bench, ate lunch early, and watched the afternoon tilt. Same trip, different city: dinner at 21:40 when the kitchen stopped at 21:00, or a palace at 16:50 when last entry was 16:00. The draft sounded dense and clever. Calendar reality did not care.

This page is about opening hours as constraints in AI day plans. Closed weekdays. Last-entry cutoffs earlier than closing time. Seasonal shutters. Kitchen last-order windows that are not the same as "dinner somewhere nice." Not ticket slots, not buffer minutes, not overall pace. Those have their own guides. Here you teach the model which doors are open on which calendar days and until which clock. Most of it works in ChatGPT, Claude, Gemini, or similar even if you never sign up. Near the end we note when a preference survey that holds these rules across revisions can beat paste-and-hope.

Also see: How to fit timed entries into an AI itinerary · How to leave open time in an AI itinerary · How to plan a trip around your pace · How to set a walking budget for an AI itinerary · How to protect arrival and recovery days · How to choose bases for a multi-city AI itinerary · How to refine an AI itinerary without starting over · How to set food rules for an AI itinerary · How to stop getting generic AI itineraries · What is Dople? · Guides · Samples (ILLUSTRATIVE) · llms.txt

What opening-hours rules are (and what they are not)

Opening hours are calendar facts. This museum is closed Mondays. That palace last-entry is 16:00 even if the building "closes" at 17:30. The hill town church closes for the season from November through March. The neighborhood kitchen stops taking food orders at 21:00 even if the bar stays loud until midnight. You write those facts into the brief so the model cannot invent a day that ignores them.

They are not timed entries. A timed entry is a ticket slot you bought for Tuesday at 11:00. Opening hours decide whether Tuesday at 11:00 is even a legal time to show up. They are not open time either. Open time is the slack between blocks so a late tram does not collapse lunch. Opening hours decide whether the place exists as an option that afternoon at all. Pace is week-level energy. Walking budget is distance. Bases are where you sleep. Opening hours are the doors.

If you only say "visit the famous museum," models often pick a random weekday and a polite morning slot. If you say "Museum X closed Mondays; last entry 16:00; do not schedule after 15:15 arrival; Kitchen Y last order 21:00; seasonal site Z closed Nov-Mar," the draft has fewer ways to strand you on a plaza with a locked door.

Why AI drafts ignore calendar reality

Models love a complete-looking day. Naming a famous museum on Monday still looks complete in prose. Training data mixes cities where museums close Monday with cities where they do not, and averages quietly win unless you lock the exception. Last entry is worse: many pages say "open until 18:00" while the ticket desk stops an hour earlier. The model copies the later number. You arrive at 17:20 and get turned away.

Seasonal hours vanish in generic briefs. A summer guidebook tone sneaks into a February trip. Hill towns, gardens, ferry docks, and outdoor ruins get treated as year-round. Kitchen hours get confused with restaurant "open" hours. A place can be open for drinks while the kitchen is dead. Food rules that say "nice dinner" without a last-order clock still let the model pin 21:45 at a kitchen that closed at 21:00.

Holidays and local quirks rarely appear unless you force them. National holidays, saint days, surprise Tuesday closures, and renovation notes will not save you if the brief only lists attraction names. Multi-city trips multiply the risk: each base has different closed-day norms, and a transfer morning into a Monday museum is a classic self-own.

A pasteable opening-hours brief

Paste this as the calendar source of truth before you ask for morning-afternoon-evening detail. Fill brackets with real closed days, last-entry times, seasonal months, and kitchen windows. Attach timed-entry, open-time, walking, pace, spend, food, lodging, and dealbreaker lines you already trust. Ask for a day-by-day that never invents a Monday museum, a post-last-entry visit, or a dinner after kitchen close.

Trip: [city / region], [N] nights, [month/season], travelers: [N]

OPENING-HOURS RULES (locked; do not schedule over these unless I reopen them):
- Closed days (must-see stops):
 - [Museum / site A]: closed [Mon / Tue / ...]; open [hours] other days
 - [Museum / site B]: closed [days]; open [hours]
 - [Market / shop cluster]: closed [days] or half-day [day]
- Last entry (separate from closing time):
 - [Site A]: last entry [e.g. 16:00]; plan to arrive by [e.g. 15:15]; do not schedule after that
 - [Site B]: last entry [time]; arrive-by [time]
- Seasonal / shuttered:
 - [Site / route / ferry / garden]: closed or limited [months]; if our dates fall inside, replace with [alt] or drop
- Kitchen / meal windows (clock, not vibe):
 - Lunch kitchens: last order roughly [e.g. 14:30] unless named exception
 - Dinner kitchens: last order roughly [e.g. 21:00 / 22:00]; bar-open is not kitchen-open
 - Named restaurants we care about: [Name] kitchen last order [time]
- Holiday / special closures to check before locking: [list dates or "flag national holidays in [country]"]

MUST NEVER:
- Do not put a closed-day stop on its closed weekday
- Do not schedule arrival after last entry, even if "closing time" is later
- Do not treat bar hours as kitchen hours
- Do not use summer hours for a winter trip (or the reverse) without saying so
- Do not invent "open late" for sites that post early last entry

TIMED / OPEN-TIME / WALKING / PACE / BASE CONSTRAINTS (paste short lines):
- Timed entries allowed: [which days / which pins] (ticket slots still need the place open)
- Open-time buffers: [paste short open-time lines]
- Walking ceiling: [paste from walking-budget brief]
- Pace notes: [paste short pace lines]
- Arrival / recovery / transfer days: light only; still respect closed days and last entry
- Overnight bases: [cities / neighborhoods]

OUTPUT I WANT:
- Day-by-day with weekday named on each day (Mon, Tue, ...)
- Each hard stop tagged with open / closed / last-entry note when relevant
- Meal pins that respect kitchen last-order windows
- Checklist: no closed-day pins; no post-last-entry arrivals; no post-kitchen dinners; seasonal shuttered sites swapped or dropped

One follow-up that catches calendar failures early: ask the model to list every hard stop with the weekday, the posted open hours it used, last entry if any, and kitchen last order for meal pins. If a museum lands on its closed day, or arrival is after last entry, or dinner is after kitchen close, the brief did not stick. Send that stretch back by name. Demand a calendar repair without regenerating the whole trip.

For shared constraint lines, lean on briefs you already trust: timed entries, open time, pace, walking budget, arrival and recovery, multi-city bases, food, and generic AI itinerary.

How opening hours sit with tickets, buffers, walking, and bases

Timed entries need the place open that day. A Tuesday 11:00 ticket is useless if you somehow shifted the whole day and the site is closed Tuesdays, or if last entry is 10:30. Use the timed entries guide to place slots on the right days, and this page to make sure those days and clocks are legal against posted hours.

Open time keeps you from racing between open doors. Opening hours decide which doors exist. A twenty-minute buffer before last entry is not optional politeness; it is how you still get in when the tram is slow. Pair with the open time guide so last-entry pins keep slack on both sides.

Walking budget caps distance. Opening hours can force a longer walk if the only open museum that day sits across town. Raise the ceiling for that day, take transit, or swap the stop. Do not accept a locked-door day because the nearer museum was closed. See the walking budget guide.

Arrival and recovery days already want light plans. They still need calendar facts. A "light" first afternoon that pins a Monday-closed museum is not light; it is a plaza nap you did not choose. Pair with the arrival and recovery guide.

Multi-city bases decide which city's closed-day norms apply on which overnight. A transfer morning into a Monday-heavy museum city needs a different plan than a full Tuesday on site. Lock bases with the multi-city bases guide, name the weekday on transfer days, and keep opening-hours rules per city, not one averaged fantasy schedule.

Food rules shape cuisine and reservation style. Kitchen windows decide whether that dinner is reachable by clock. Spend caps still matter when a closed door panics you into a taxi. Pace decides how many hard mornings you want; opening hours decide which mornings are available.

Red flags in AI output

Museum Monday. Or Tuesday, or whichever local closed day the draft ignored. If the weekday is wrong for a must-see, the day is wrong. Move the stop or change the day.

Arrival after last entry. Closing time said 18:00. Last entry was 16:30. The draft put you at 17:10. That is a locked ticket desk with a pretty facade.

Seasonal shutter treated as open. February visit to a site that sleeps until April. Ferry schedule from July pasted into November. Demand a seasonal check or a replacement.

Kitchen closed, bar open. Dinner pin at 21:50 when last order was 21:00. Or lunch at 15:10 in a city where kitchens pause mid-afternoon. Bar hours are not meal hours.

Holiday blindness. Draft assumes a normal weekday on a national holiday when museums and markets run Sunday hours or shut entirely. Ask for a holiday flag before you lock tickets.

Same hours every day. A draft that never varies open times by weekday, even when the site posts late Fridays and early Sundays. Force weekday-specific hours for must-sees.

Transfer day plus closed-day pin. You arrive Monday midday and the only plan is a Monday-closed palace. Build an open alternative for that calendar.

Rules live only in chat. By revision three the model forgets Monday closures and last entry. Paste the opening-hours brief into every refine until weekdays and last-entry tags stay visible.

Refine without restarting the whole trip

A closed-door day rarely needs a full regenerate. Regenerating often reprints the same Monday museum with a new cafe name and the same kitchen mistake at dinner. Switch to an in-place refine.

Name the broken pin first. "Move Museum A off Monday; it is closed Mondays. Put it on Tuesday or Wednesday. Keep lodging and the evening show." Specific beats "fix the morning."

Separate last entry from closing time. Ask for arrive-by times on every last-entry site. If arrive-by is after last entry, move earlier or drop it that day.

Swap seasonal ghosts. If a site is shuttered for your months, demand a named replacement nearby, not a vague "explore the area."

Fix kitchen windows without reopening the whole food brief. "Dinner at Restaurant Y must start by 20:30; kitchen last order 21:00. Do not move other locked meals."

Restate timed-entry and open-time lines if tickets or buffers caused the scramble around the closed door. The refine guide has the keep/change/drop pattern. Add one opening-hours line: "Keep locked stops [A, B]; move [C] off closed day [weekday]; arrive by [time] for last entry at [D]; dinner kitchen last order [time]; do not add a new must-see that ignores posted hours."

When two travelers argue about whether to "just go and see," settle with the posted hours in hand. The travelers disagree guide helps pick which stops stay hard when the calendar forces a swap.

How this sits with the other briefs

This page is calendar reality: closed days, last entry, seasonal shutters, kitchen clocks. The others keep tickets, slack, walks, and overnight maps from fighting it.

Timed entries place ticket slots. Opening hours decide whether those slots land on an open day before last entry.

Open time is slack between blocks. Opening hours are whether the next block's door is open at all; buffers still matter before last entry.

Pace sets week-level density. Opening hours constrain which mornings and evenings are even available.

Walking budget caps distance when the only open site that day sits farther away.

Arrival and recovery keep first and last days light. Opening hours still apply so light days do not pin closed doors.

Multi-city bases lock where you sleep and which city's closed-day norms apply on transfer mornings.

Food shapes meals. Kitchen last-order windows from this page protect the clock around them.

Daily spend budget caps panic taxis and tourist-trap dinners after a closed door.

Lodging and dealbreakers keep hotel and hard-no rules from adding surprise pins that ignore local hours. Generic AI itinerary work is the fuller stack before draft one. Refine is the edit loop after a day still ignores closed days or last entry.

When a preference survey and Trip Project path helps

If you like prompt work and you will keep pasting the opening-hours brief yourselves, a general assistant plus the calendar rules above is enough for many trips.

The structured path starts to matter when chat keeps putting you on closed weekdays after you locked the brief. Or when last-entry times, kitchen windows, timed tickets, and open-time gaps live in one messy thread and drift by revision four. Or when you want a researched day-by-day that holds those constraints without re-teaching the model every session.

One option with that shape is Dople (dopletech.com). A short preference survey can capture opening-hours style constraints alongside pace, food, walking, and lodging, you compare trip options shaped by it, and you can buy a Trip Project: a researched day-by-day meant to hold those preferences across revisions so you are not stuck reprinting Monday museums. Account and survey start are free. Full itineraries are paid. More detail: Dople vs ChatGPT and What is Dople?.

Public samples show ILLUSTRATIVE day-by-day shapes only. They are not live bookings and not free Trip Projects.

If the survey path sounds useful, sign up once when you are ready.

What are opening-hours rules in an AI itinerary?

Opening-hours rules are calendar facts the model must respect: closed weekdays, last-entry cutoffs, seasonal closures, and kitchen last-order windows. Without them, drafts put you at locked doors or after the kitchen closes. Timed entries are ticket slots; open time is slack; opening hours are whether the place is open that day and until what clock (see timed entries and open time).

How is this different from timed entries or open time?

Timed entries are slots you buy. Open time is buffer minutes between blocks. Opening hours are posted calendar reality: closed days, last entry earlier than closing time, seasonal shutters, kitchen last order. You usually need all three so a legal ticket still lands on an open day with slack before last entry.

What belongs in an opening-hours brief?

Closed days per must-see stop, last-entry times separate from closing times, seasonal closed months, kitchen last-order windows, and must-never lines that forbid Monday museum pins or dinner after kitchen close. Paste timed-entry, open-time, walking, and base rules alongside it. Pair with food rules when cuisine constraints and kitchen clocks both matter.

Should I regenerate when the draft ignores closed days?

No. Regenerating often repeats the same Monday museum with a new lunch name. Refine in place: move the closed-day stop, swap the early kitchen, restate last entry, and keep locked pins (see the refine guide).

Are Dople samples real bookings?

No. Samples are ILLUSTRATIVE examples of itinerary shape. Real Trip Projects are researched per traveler after the preference survey. More context: Personalized travel itineraries FAQ.