Make life easierGuide

Plan small-gathering table groups from agreed preferences

Keep every guest accounted for, honor explicit together/apart rules and return a clear no-fit when needed.

Illustrative resourceNot tool-testedOrganizers making a small table-group plan from an authorized guest list and explicit preferences

What you’re working towards

A complete guest-to-table list, a constraint ledger and a readable two-table locator. The example assigns groups only, not clockwise seats, a measured venue layout or social compatibility.

Download playbook .md

Your choice also saves this guide’s current worked-example checkmarks. Only this browser; no account or sync. View saved guides

Saving needs JavaScript. Reading and downloading this guide does not save it.

Included example · illustrative CSV file

Download the example table assignments

Eight fictional guest IDs, each assigned exactly once, with matching table IDs and labels.

Not tool-tested. Review and adapt before using it for a real project. Your browser may download this file when you choose View example file.

View example file Download example

Synthetic table-group locator. Cedar lists guests A, B, F and H. Pine lists guests C, D, E and G. Each table has four of its four places assigned. This is a group list, not a clockwise seat order or a venue map. — A readable locator for the supplied fictional groups. Confirm real preferences, capacities, physical access and guest-list sharing separately.

Source files to work with

Download the complete synthetic brief
All guests, capacities, preferences and explicit exclusions before the answer; no private relationship data.
Download the structured brief
The same complete supplied guest IDs, table capacities, anchor and together/apart pairs.
Download the checked constraint ledger
Ten explicit checks covering the anchor, all pairs, both capacities and complete guest coverage.
Download the explanation and no-fit examples
Full assignment reasoning and narrow enumeration result, without real-event or fairness claims.
Download a blank organizer brief
Reusable unfilled input and review checklist; real guest use and sharing require organizer decisions.
Download the editable table locator
Self-contained original vector artwork listing the exact groups, not venue geometry.
Download the table locator image
1200 × 720 actual render with four guest IDs under each table name.

Before you begin

A confirmed complete guest list you have permission to use
Named tables and confirmed capacities
Only the organizer-supplied together/apart preferences and explicit table anchors
Separate confirmation of real room/access needs and permission to share the final guest list

Possible tools

A text AI assistantUser-chosen paste-text interface · required
Access and plan limits unverified
No named AI product, plan, free route, printer or receiving application was verified. Use tools you already have access to; review actual guest permissions and venue needs separately.
An independent constraint checkManual ledger or bounded local enumerator · required
Access and plan limits unverified
No named AI product, plan, free route, printer or receiving application was verified. Use tools you already have access to; review actual guest permissions and venue needs separately.
A text or image viewerUser-chosen local viewer · suggested
Access and plan limits unverified
No named AI product, plan, free route, printer or receiving application was verified. Use tools you already have access to; review actual guest permissions and venue needs separately.

Use the sample first. Remove private information before taking a prompt into another tool.

Make a start

One step at a time.

4 steps

Start with the synthetic worked example or other non-sensitive inputs. Run one prompt at a time in your chosen AI tool. Replace any bracketed placeholders with the requested material, check each reply, and supply the source or earlier outputs that step needs. Paste them again if the tool lacks that context.

Set up the worked example

Open: Open a fresh text-AI chat and a plain-text editor. Keep the guide open to check every ID and rule. You can make a readable table locator as text; image generation or a venue-mapping tool is not required.
Add the sample: Download and open seating-source.md, or copy all of “The starting point” (Synthetic input), into step 1’s BRIEF field. In step 2 paste the complete checked brief. Save the proposed groups as my-table-groups.txt and reopen them; step 3 needs both the unchanged original brief and that saved assignment in its two fields. After checking every rule, step 4 needs the checked assignment plus this sample display instruction: “Use only the fictional IDs A–H and the exact Cedar/Pine table labels. This is a private practice locator; do not publish or send it.” Save its full text as my-table-locator.txt and reopen it for comparison. “The intended result” (Illustrative output), CSV ledger and existing PNG/SVG are comparison answers. The optional advanced exercise belongs to this sample chat: paste the full original brief and saved plan, then explicitly state a fictional practice change and which other inputs stay fixed. In a new chat supply all of that again; do not infer organizer approval for real changes.

Start with the synthetic example, then adapt each prompt to your task.

Collect the actual rules without asking for personal explanations

Use guest IDs while planning. Confirm the complete list, capacities and explicit preferences. The example gives eight fictional adults and two four-place tables; it supplies no relationship history.

Your starting prompt

Organize this authorized table-group brief. Preserve every guest ID, table label/capacity, explicit guest-to-table anchor and together/apart rule. Confirm whether the list and preferences are complete. Flag duplicate/unknown IDs, invalid capacities or missing confirmations. Do not ask for reasons for preferences, infer relationships or use personal characteristics to decide compatibility. Do not invent room/access constraints or a seat order.

BRIEF:
[paste]

Expect: A complete list of eight guest IDs, two table capacities, A’s Cedar anchor, four together pairs and two apart pairs.
Check: The source explicitly confirms the full guest list and supplied preferences. Check A/B, C/D, E/G and F/H together; A/C and A/E apart. No reason for a preference, personal trait, chair position or physical access fact is supplied.
If it differs: If a guest or rule is missing, paste the corresponding source sentence and request a corrected brief. Treat an unconfirmed list or preference set as a review hold rather than “no restrictions”. Do not invent a social explanation.

Ask for complete groups and keep no-fit visible

The example places A/B/F/H at Cedar and C/D/E/G at Pine. Every guest has one table. If a valid source has no solution, the organizer must decide which input to change.

Your starting prompt

Produce a guest-to-table assignment satisfying all supplied rules. Assign every guest exactly once, keep each table within capacity, and honor every anchor and together/apart pair. Sort guest IDs within each named table for readability. Return only groups, not clockwise seat positions. If no valid assignment exists, return no fit and identify the conflicting supplied constraints without relaxing them or dropping guests. Treat missing information as a review hold, distinct from a proven no-fit.

VALIDATED BRIEF:
[paste]

Expect: One complete table assignment or a supported no-fit finding. Save the assignment in my-table-groups.txt.
Check: For this sample Cedar is A, B, F, H; Pine is C, D, E, G. Count four at each table and eight different IDs overall. A table group does not specify a seat order.
If it differs: If the answer drops someone, exceeds capacity or weakens a rule, name the failure and request a new complete assignment from the unchanged brief. A failed guess alone is not proof of no fit; ask for a constraint-based reason or leave feasibility unverified.

Check every guest and every preference independently

Use the constraint ledger to verify coverage, capacities, anchor and all six pair rules. For this fixed eight-person example, all 70 four-person Cedar subsets were checked and only one satisfies the full brief.

Your starting prompt

Audit the assignment against the original brief. List each guest exactly once with their table; count each table; check every anchor and each together/apart pair individually. Show missing, extra or duplicate guests explicitly. For this small fixed example, independently enumerate all four-person Cedar subsets with the complementary Pine group and compare valid results; do not claim general optimality or fairness. Test a ninth guest at the same capacity, a five-person together group and unconfirmed preferences. Keep local arithmetic/constraint checks separate from actual organizer, venue or consumer-AI testing.

ORIGINAL BRIEF:
[paste]

ASSIGNMENT:
[paste]

Expect: A guest-by-guest audit and a line for each anchor, pair, capacity and coverage check, using the reopened assignment.
Check: Check all ten ledger items yourself. To cross-check uniqueness without code: A/B must occupy Cedar; C/D and E/G must fill Pine because C and E are apart from A; F/H therefore occupy Cedar. The prompt also requests an exhaustive check of 70 four-person Cedar subsets. If no enumerator is available, label that enumeration not run; the forced-placement reasoning is a separate manual check, not an enumeration receipt. A ninth guest exceeds eight places; five required-together guests exceed either four-place table.
If it differs: Keep tests separate from the original brief. Unconfirmed preferences should hold, not count as an impossible complete model. Correct any failed assignment in a new saved revision and recheck every rule, since moving one guest can break several pairs.

Make a clear table locator after the organizer reviews it

The downloadable image lists the four guest IDs under each table name. It deliberately avoids drawing seat positions or a room map. Confirm real sharing and access decisions before printing a real version.

Your starting prompt

Create a simple readable table locator from the reviewed assignment. Use the exact table labels and guest names/IDs the organizer has authorized for display, with each guest once. Keep the list alphabetical unless a different display order is explicitly supplied. Label the result as table groups, not seat order or a measured venue map. Supply a full text equivalent. Do not add reasons for pair rules, inferred relationships, personal characteristics or unapproved guest data.

REVIEWED ASSIGNMENT AND DISPLAY PERMISSION:
[paste]

Expect: A simple locator listing exact table labels and authorized fictional IDs, plus a complete text equivalent. Save and reopen my-table-locator.txt.
Check: Read the reopened locator against my-table-groups.txt: Cedar A/B/F/H and Pine C/D/E/G, each ID once, sorted within its table. If an image was genuinely created, inspect its visible labels as well; a claim of a file is not a saved image.
If it differs: If a label is missing or an image cannot be created/read, keep a corrected full text locator. Do not use an invented room map or chair order. Real use requires organizer review, current capacities/preferences, venue/access checks and permission for the intended audience before sharing.

Optional: improve the result

Show exactly what changes after an organizer decision

Keep the original brief. Recalculate only after an explicit guest, table or preference change, and explain any moved guests without inventing a social rationale.

Advanced prompt

Compare an updated table-group brief with the unchanged original. Apply only the organizer’s explicit changes below. Validate the new source, find a complete assignment or a clear no-fit, and check every original/changed constraint. Return a guest-level change list showing old table, new table and the supplied constraint change that required reconsideration. Do not minimize moves unless asked, claim fairness, infer relationship reasons or overwrite the original plan.

ORIGINAL BRIEF AND PLAN:
[paste]

EXPLICIT APPROVED CHANGES:
[paste]

Optional: use your own input

Propose your own table groups

Use fictional IDs or an authorized non-sensitive adult guest list. Do not include relationship explanations. Start a new chat with the completed prompt below, optionally using seating-blank-brief.md as a local source worksheet. Check every ID, table and rule in the response, then send:
“From MY COMPLETE CHECKED BRIEF, propose table groups with each guest exactly once, no table above its capacity, all explicit anchors honored and every together/apart rule obeyed. Sort IDs within each named table. Do not infer compatibility or seat order. If you find no arrangement, show a sound constraint conflict or exhaustive search before calling it no fit; otherwise say the search is unverified. Missing confirmation remains a hold. COMPLETE CHECKED BRIEF: [paste all guests, tables, constraints and confirmation states].”
Count guests and compare the proposal with the source. Save it as my-table-groups.txt and reopen it.
“Audit MY REOPENED ASSIGNMENT against MY ORIGINAL BRIEF. Make a coverage list for every guest and a separate check for every capacity, anchor and together/apart rule. Report missing, extra or duplicate guests. Test an over-capacity case and an unconfirmed-input case on separate copies; do not alter the original. Do not assume my case has the sample’s counts, unique answer or two tables. ORIGINAL BRIEF AND REOPENED ASSIGNMENT: [paste both].”
Review each check yourself; correct and recheck a new revision if needed. Obtain the organizer’s review and explicit display-label/audience permission separately. If permission is pending, keep a private ID-only draft and label it pending.
“Create a readable text table locator from MY REVIEWED GROUPS and MY DISPLAY PERMISSION. Use only authorized labels, each guest once, alphabetically within each table unless my permission specifies another order. Include a full text equivalent for any separately requested image. No preference reasons, inferred traits, chair positions or venue geometry. Do not send or publish it. REVIEWED GROUPS, EXACT DISPLAY LABELS, INTENDED AUDIENCE AND PERMISSION STATE: [paste].”
Save as my-table-locator.txt and reopen it; compare every visible label with the checked group list. If a genuine image is created, save/open it too and compare it with the text; otherwise the text locator is sufficient. Hold any sharing until permission is confirmed. The original advanced exercise remains sample-scoped; an own-input revision needs the complete original brief/plan, explicit organizer-approved changes and confirmed unchanged inputs. Recheck actual venue/access constraints separately before using groups.

Your own-input prompt

Help me propose ordinary table groups from supplied rules only. First organize the complete brief: unique guest IDs, table IDs/labels and positive integer capacities, explicit table anchors, together/apart pairs or groups, confirmation that the guest list and preferences are complete, and display permission state. Validate unknown or duplicate IDs, conflicting rules and missing confirmations. Use no personal characteristics or relationship explanations, infer no other preferences, and do not decide social compatibility, fairness, physical access or seat order. Stop for my review before assigning groups. Later we will propose groups, save/reopen them, check every rule and prepare a private locator with only permitted labels. Do not invite, contact, publish or assign real seating.

MY COMPLETE AUTHORIZED BRIEF
[Give all guest IDs; every table label/capacity; every explicit anchor/together/apart rule, or an explicit confirmed none; whether the list/preferences and capacities are confirmed; who reviews the proposal; permitted display labels and intended audience, or permission pending. Keep venue/access checks separate and flag unresolved inputs.]

Check your own result

Every guest in your own list appears exactly once; no sample IDs, table labels or counts remain by accident.
Each capacity, anchor and together/apart rule is checked individually; no rule is relaxed to produce an answer.
Unconfirmed inputs, a search that found no answer and a proven no-fit result remain different outcomes.
The saved locator matches the checked groups and uses only authorized display labels. It does not establish seat order, fairness, compatibility or venue accessibility.
Organizer review, actual venue/access needs and permission for the specific sharing audience are confirmed before real use.

Show, then try

See the shape of the result.

This input is synthetic. The output is a written illustration of the intended result, not a verified AI-tool run.

The starting point

Synthetic input

# Complete synthetic table-group brief

Eight fictional adults use IDs A, B, C, D, E, F, G and H. There are no real names or relationship explanations. The organizer confirms this is the complete list and that no preferences beyond those below are supplied.

- Cedar has capacity 4.
- Pine has capacity 4.
- A must be at Cedar.
- A and B stay together.
- C and D stay together.
- E and G stay together.
- F and H stay together.
- A and C must be at different tables.
- A and E must be at different tables.

Assign every guest exactly once and never exceed either capacity. Honor only these supplied rules. Show guest IDs alphabetically within each named table. These are table groups, not clockwise seat positions or a measured venue map.

Do not ask for or invent reasons for the pair rules. Do not infer friendship, compatibility, social rank, age or other personal characteristics. No room geometry, physical access or real venue constraints are supplied. The use of fictional adults is only a data-provenance statement, not a factor in the grouping.

If the source is valid but all assignments violate a rule, return no fit and ask the organizer what they wish to change. Do not silently remove a guest, increase capacity or relax a rule. Hold missing/unconfirmed guest or preference information for review; an absent rule is not permission to invent a replacement.

The result is an illustrative planning answer. The organizer must confirm real guest preferences, table capacities, physical room/access needs and permission to share any real guest list before using it. No invitation, actual event or consumer AI execution is part of this fixture.

Copies text only. For image or file tasks, use the source files named in the steps too.

Download source .txt

The intended result

Illustrative output

# One complete table assignment

- Cedar: A, B, F, H.
- Pine: C, D, E, G.

Every fictional guest appears exactly once. Both tables contain four guests, matching their supplied capacities. A is at Cedar. Each together pair shares a table, and A is apart from C and E.

## Why this fits

A's anchor puts A at Cedar, so B follows. C must be apart from A, so C and its together partner D go to Pine. E must also be apart from A, so E and its partner G fill Pine. F and H stay together and occupy the two remaining Cedar places.

A local exhaustive enumeration considers all 70 choices of four guests for Cedar, with the other four at Pine. Exactly one labelled-table partition satisfies this complete fixed input. This does not claim that real events have a unique, best or fair arrangement. No clockwise seat order is chosen.

## If the source changes

A ninth guest with the same two four-place tables has no fit. A required group of five cannot all fit at either four-place table, even with other pair restrictions removed. Missing/unconfirmed preferences remain a review hold; they are not the same as a valid source with no solution.

Preserve the original brief, request the organizer's decision and recompute after any explicit change. Confirm actual preferences, capacities, physical layout/access and guest-list sharing before using a real arrangement. No real event, invitation, compatibility assessment, consumer-AI workflow or time saving was tested.

Download result .txt

You’re the final check

Check the worked example.

These checks use the worked example and its particular facts. With your own material, check your actual names, numbers, sources and rules; do not reuse the sample’s values as the answer.

Cedar contains A, B, F and H; Pine contains C, D, E and G.All eight fictional IDs appear once, and both table counts equal four.A is at Cedar; A/B, C/D, E/G and F/H stay together.A is apart from both C and E. No other preference is inferred.All 70 capacity-fitting Cedar subsets yield exactly one valid labelled-table partition for this fixed source.A ninth guest or a five-person required group has no fit at the unchanged capacities; unconfirmed input is a distinct hold.The diagram shows group membership only and exactly matches the CSV. Real venue/access and guest-list sharing remain unverified.

Unsaved checks stay in this tab and reset when you reload. If you choose Save guide, its current example checkmarks are kept in this browser when storage is available.

A few things to watch for.

Dropping a guest or increasing a table capacity to make a plan appear complete.
Silently weakening an apart/together rule or inferring a reason behind it.
Treating missing preferences as confirmation that none exist.
Calling one fixture solution the fairest, best or socially compatible arrangement.
Drawing chair positions or venue geometry when only table groups were supplied.
Publishing real names or sensitive relationship notes without the appropriate permission.