Counts show all available guides. Filters combine to narrow your results.
Make life easierGuide
Compare furniture layouts using supplied measurements
Check three room arrangements against fixed footprints and personal keep-clear patches before drawing conclusions from a picture.
Illustrative resourceNot tool-testedAnyone comparing furniture positions from a small set of measured rectangular footprints
What you’re working towards
A coordinate plan, equally scaled three-option visual and text comparison showing one arrangement that satisfies the supplied mock rules and two that fail exact geometric checks.
All furniture coordinates, exact intersections, boundary failures, reading order, legend, limits and the 410 cm room-width perturbation.
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 new text-AI chat, a calculator and a plain-text editor. Download layout-measurements.md. You can calculate from its numbers without sharing a room photo. To finish the visual checks you also need a precise SVG drawing/export tool and image viewer; if unavailable, retain the coordinate/text comparison and mark the drawing checks pending. A footprint means the rectangle occupied by an item in this plan.
Add the sample
In step 1 replace [paste layout-measurements.md or your own verified measurements] with the complete downloaded measurement brief, matching ‘The starting point’ (Synthetic input). Keep the optional 410 cm change separate from the base 480 cm room. In step 2 paste the complete original brief and checked measurement summary; in step 3 paste that source and the checked coordinate results. Save and reopen your new coordinate table, SVG and text before step 4; paste the original brief and reopened CSV/text under their headings. Attach your exported PNG only if image inspection is supported, or state that image review is unavailable and supply source for a source-only audit. ‘The intended result’ (Illustrative output) and reference files are a comparison key after your attempt.
Start with the synthetic example, then adapt each prompt to your task.
01
Lock the measurements and coordinate rules
Start with the source brief and the three supplied positions. Check units, origin, orientation and the meaning of each footprint before using a drawing. An omitted feature remains unknown.
Your starting prompt
Read the supplied measurement brief. Extract the room bounds, fixed furniture dimensions, lower-left positions for each candidate option, and every user-supplied keep-clear patch. State the unit, origin, x/y directions, rotation rule and edge-contact rule. Check for missing dimensions, mixed units, nonpositive sizes, duplicate items and keep-clear patches outside the room. Ask for ambiguous values rather than guessing from a photograph. Treat these as arbitrary personal constraints, not safety or accessibility standards. Do not move, resize, rotate or omit an item to make an option fit.
SOURCE BRIEF:
[paste layout-measurements.md or your own verified measurements]
Expect
A locked measurement summary: a 480 × 360 cm room, four fixed furniture sizes, two personal keep-clear patches and all twelve placements across A, B and C.
Check
The origin is lower-left, x runs right and y runs up. A position is an item’s lower-left corner. No rotation is allowed; touching edges is allowed but an intersection with positive area is not. K1 and K2 are arbitrary exercise preferences, not safety clearances.
If it differs
If it mixes units, changes a size/position or guesses a door, ask it to restore the source and list the unknown separately. Missing or invalid measurements must be resolved before an overall pass.
02
Calculate every extent and intersection
Calculate before drawing. The upper-right corner is x plus width, y plus depth. Check all six furniture pairs and all eight furniture-to-zone pairs in every supplied option.
Your starting prompt
Using the locked measurements, return a coordinate table with option, item ID, item name, x, y, width, depth, right edge and top edge, all in cm. Recalculate all four bounds per item. For every furniture pair and every furniture/keep-clear pair, use positive-area rectangle intersection: max(0,min(right edges)-max(left edges)) multiplied by max(0,min(top edges)-max(bottom edges)). List each nonzero intersection with its exact x/y interval and area. List each out-of-room extent and area separately. Touching at an edge or corner is allowed only because this fixture explicitly says so. Mark an option as satisfying the mock rules only if the inputs are complete and valid and every required check passes. Do not infer comfort, walkability, installation suitability or compliance.
LOCKED MEASUREMENTS:
[paste]
Expect
A coordinate table with right edge = x + width and top edge = y + depth, plus checks of all six furniture pairs and eight furniture/zone pairs per option.
Check
A meets the mock rules. B’s sofa/table overlap is x390–420 and y100–125: 30 × 25 = 750 cm². C’s desk ends at x520: its outside strip is 40 × 70 = 2,800 cm², and its K2 intersection is 100 × 70 = 7,000 cm². Check these with your calculator.
If it differs
If a failure is hidden by moving or resizing an item, restore its original coordinates. Ask for the exact shared x and y intervals and multiply their positive lengths; edge contact gives area zero. Keep boundary and zone failures separate.
03
Draw each option at the same scale
Use the coordinate table as the source for precise rectangles. Keep failed areas visible, label the fixed furniture and reserved patches, and put the full result in text alongside the image.
Your starting prompt
Create a self-contained SVG plan comparing the three checked options at one identical x/y scale. For this fixture use 0.9 image pixels per cm. In SVG, convert the upward room y axis to the downward image axis using image_y = origin_y - scale*(y+depth), with image_x = origin_x + scale*x. Set drawn width = scale*width and height = scale*depth. Show every whole room, all four unchanged furniture rectangles and both keep-clear zones in each panel. Show out-of-room parts rather than clipping them. Label axes, cm units, origin convention, furniture IDs/sizes, personal zones and each failed intersection. Use labels and patterns as well as color. Add a short image description and a separate complete text equivalent containing every coordinate, failure and limitation. Do not add inferred doors, circulation clearances or real-world suitability claims. Add no scripts, links, external images, fonts or network resources. Do not claim that a file was rendered or viewed unless it was.
SOURCE AND CHECKED COORDINATES:
[paste]
Expect
Complete SVG source for three equally scaled plans, with labels/patterns for failures, and a full text comparison. Save new files as my-layout-plan.csv, my-layout-comparison.svg and my-layout-notes.txt, excluding chat commentary/code fences.
Check
Each direction uses 0.9 pixels per cm. SVG y runs down, so its top edge uses origin_y − 0.9 × (y + depth). All rooms, zones and full furniture rectangles stay visible, including C’s desk beyond the wall. The CSV has twelve rows and no moved or resized pieces.
If it differs
If SVG source or a table is missing, request the complete result. A text AI cannot promise a PNG: open the saved SVG in a capable viewer and export my-layout-comparison.png only if supported. Otherwise mark raster checks pending. Reopen the saved source/table/text before auditing.
04
Open the image and cross-check the result
Inspect the exported raster, then independently recalculate its rectangles against the source. A polished plan can still hide an overlap or silently change a dimension. The reference files are a comparison key after your attempt.
Your starting prompt
Audit the exported image, coordinate CSV and text alternative against the original measurement brief. Independently recompute each right/top edge, furniture overlap, zone overlap and boundary overrun. Check every rectangle uses the same scale and the y conversion is correct. For this fixture verify A satisfies the mock rules; B overlaps S/T over x390–420,y100–125 for 750 cm²; C overlaps D/K2 over x380–480,y290–360 for 7,000 cm² and extends to x520, giving a 40×70=2,800 cm² outside strip. Open the actual PNG and check all labels, full footprints, the K2 tag beneath C’s desk, legend and notes are legible and unclipped. Keep source/code checks, actual image observations and unperformed consumer-AI/browser/real-room tests separate. Report changed or unsupported claims. Keep static image, CSV and text updates together.
ORIGINAL BRIEF:
[paste]
OUTPUT CSV AND TEXT:
[paste]
EXPORTED IMAGE:
[attach or open PNG]
Expect
A recalculated comparison and, if available, observations from the actual saved/exported image. Source checks, image observations and unperformed real-room tests are separate.
Check
Verify 750, 7,000 and 2,800 cm² again against the original brief. Open the PNG yourself and check all three panels, furniture letters, zone labels, axes, notes and the K2 label below C’s desk are readable. A satisfying these limited rules proves no real-room safety, comfort or accessibility.
If it differs
For any mismatch, fix the coordinate table, drawing and notes together, save new revisions and reopen them to recheck. Missing image access means no visual pass. Do not apply the sample’s conclusions to a real space with missing physical features.
Optional: improve the result
Perturb one measurement and invalidate stale conclusions
Change only the room width to 410 cm and keep everything else fixed. K2 then lies partly outside the room, so the source must be resolved before any new recommendation. The exercise exposes a failure that a furniture-only check misses.
Advanced prompt
Run a separate sensitivity check using the original fixture, changing only room width from 480 cm to 410 cm. Keep the origin, every furniture position and footprint, and both zone positions/sizes unchanged. First validate the zones: K2 still ends at x480 and extends 70 cm outside the revised room. Mark the source geometry invalid and withhold an overall pass/recommendation until the user resolves K2; do not shrink it. Then report conditional geometric effects: the sofa ends at x420 in A/B/C, producing a 10×85=850 cm² outside strip; B’s side table ends at x440, producing 30×50=1,500 cm²; C’s desk ends at x520, producing 110×70=7,700 cm². Recompute all other checks, noting pair and unchanged-zone intersections remain unchanged. Label this as a separate hypothetical revision and preserve the original result. Explain what new input is required before redrawing a valid revised plan.
ORIGINAL MEASUREMENTS AND OUTPUT:
[paste]
Optional: use your own input+
Compare your own measured arrangements
Use your own non-sensitive measurements, without an address or private room photos. This workflow compares supplied rectangular footprints; it is not a real-room approval. Start a fresh chat with the prompt below. After the source check, send: “Using my locked measurements, list each candidate’s item ID, x, y, width, depth, right and top edges in my single chosen unit. Check every furniture pair, every furniture/zone pair and every room boundary. Show exact intersection intervals and areas using max(0,min(right edges)−max(left edges)) times the corresponding y length. Apply my stated contact rule. Preserve all positions/sizes and withhold an overall pass if source geometry is invalid.” “Draw my checked candidates as complete self-contained SVG at one stated equal x/y scale. Convert the lower-left upward y coordinates correctly into SVG’s downward y coordinates. Include every room, item, personal zone and full out-of-bounds portion, with labels/patterns as well as color. Return a complete coordinate CSV and text equivalent with every result and limitation. No scripts, links, external fonts/images or inferred room features. State what has actually been rendered or viewed.” Save only the full SVG as my-layout-comparison.svg, the coordinate CSV as my-layout-plan.csv and the text as my-layout-notes.txt, using fresh filenames. Reopen the saved files. If a viewer/export tool is available, view the SVG and export a PNG; otherwise leave those checks pending. “Audit my original measurements against my saved CSV, SVG/text and actual exported image if available. Independently recompute extents, all intersections and boundary failures. Check equal scales, y conversion, unchanged footprints, readable labels and visible failures. Separate source calculations from actual image observations and unperformed real-room checks. ORIGINAL BRIEF: [paste]. SAVED CSV, SVG AND TEXT: [paste complete files]. IMAGE: [attach only if supported, otherwise state unavailable]. MY ACTUAL VIEWING OBSERVATIONS: [paste or say none].” Use a calculator to check the disputed numbers yourself. Save and reopen corrections to all affected files before repeating the audit. Your measurements replace every sample size, position, scale and conclusion. The original advanced 410 cm prompt is sample-only; for your own sensitivity check supply the exact changed measurement and revalidate the unchanged zones before drawing conclusions. Real use still needs remeasurement and omitted features/use requirements resolved separately.
Your own-input prompt
Help me compare supplied rectangular furniture layouts using only MY MEASUREMENTS. First extract the room bounds, fixed item sizes, every candidate position, personal keep-clear zones, one common unit, origin/axis directions, rotation rule and edge-contact rule. Normalize to explicit lower-left positions with x right and y up only after confirming any ambiguity. Check missing values, nonpositive dimensions, mixed units, duplicate/omitted items and invalid zones. Ask rather than guessing from a photo or moving/resizing items. Wait for my review before calculation. This is a limited geometry comparison; do not infer comfort, circulation, safety, accessibility or compliance, and do not arrange purchases or physical changes.
MY MEASUREMENTS
[Paste room width/depth and units; each item’s ID/width/depth; each candidate’s item IDs and positions; personal zone IDs/positions/sizes or explicitly none; measurement uncertainty; coordinate/rotation/contact rules. State known omitted features and the exact comparison question.]
Check your own result
All source items, dimensions and positions remain fixed unless you explicitly revise the input; units and coordinate rules agree.
Every pair, zone and boundary is checked. Invalid source geometry blocks an overall pass; failures are not cropped from the drawing.
All panels use the same x/y scale and the saved CSV, image and text describe the same revision.
A geometric result is limited to your supplied rectangles and rules; no real-world suitability or safety conclusion follows.
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.
01 / The starting point
Synthetic input
SUPPLIED MEASUREMENTS AND OPTIONS
Original fictional teaching fixture. No real home, customer or manufacturer is represented. All measurements below are supplied, exact exercise values, not estimates inferred from a photograph.
Coordinate convention
- Unit: centimetres (cm) throughout.
- Room: one rectangle, width 480 cm and height 360 cm in plan.
- Origin (0,0): the room’s lower-left corner in the plan. x increases right; y increases up.
- Every position is the lower-left corner of the rectangular footprint. Width runs along x; depth runs along y. Rotation is not allowed in the supplied options.
- A footprint includes its stated width and depth exactly. No added chair, drawer, door, cushion or circulation allowance is supplied.
- Edge or corner contact counts as no overlap for this exercise. A positive-area intersection counts as overlap. Footprints may touch the room boundary.
Fixed furniture
| ID | Item | Width (cm) | Depth (cm) |
| --- | --- | ---: | ---: |
| D | Desk | 140 | 70 |
| S | Sofa | 180 | 85 |
| B | Bookcase | 100 | 35 |
| T | Side table | 50 | 50 |
Two user-supplied keep-clear patches
These are arbitrary empty patches chosen for this mock exercise. They are not safety distances, travel paths, door swings or regulatory requirements.
| Zone | x (cm) | y (cm) | Width (cm) | Depth (cm) |
| --- | ---: | ---: | ---: | ---: |
| K1 | 0 | 0 | 90 | 100 |
| K2 | 360 | 260 | 120 | 100 |
Supplied candidate positions
| Option | Item ID | x (cm) | y (cm) |
| --- | --- | ---: | ---: |
| A | D | 120 | 260 |
| A | S | 240 | 40 |
| A | B | 40 | 180 |
| A | T | 170 | 140 |
| B | D | 120 | 260 |
| B | S | 240 | 40 |
| B | B | 40 | 180 |
| B | T | 390 | 100 |
| C | D | 380 | 290 |
| C | S | 240 | 40 |
| C | B | 40 | 180 |
| C | T | 170 | 140 |
Decision rules and requested result
Compare only these three positions without silently moving or resizing any item. Each option must include all four furniture items at their fixed sizes, keep them entirely within the room, avoid positive-area furniture overlaps and leave both keep-clear patches free of furniture. Report the coordinates and area of every failure. Make equally scaled plans, a coordinate CSV and a complete text equivalent.
No preference beyond these mock geometric rules has been supplied. Do not rank visual appeal, comfort or usability. No doors, openings, windows, sockets, heaters, wall thicknesses, vertical dimensions, installation requirements, movement routes or safety constraints are supplied. This exercise cannot establish real-world suitability, accessibility, building or fire compliance. Re-measure and obtain any relevant qualified review before using a real plan.
Optional perturbation
For a separate sensitivity check, change only the room width from 480 cm to 410 cm. Keep the origin, all furniture positions, furniture sizes and both zone positions/sizes unchanged. Identify invalid source geometry before making any new layout recommendation. Do not silently shrink or move K2.
Copies text only. For image or file tasks, use the source files named in the steps too.
ILLUSTRATIVE LAYOUT COMPARISON AND TEXT ALTERNATIVE
This is an authored reference output for a fictional geometry exercise. No consumer AI layout workflow or real room has been tested. The plans and text are static files; editing the source will not update them automatically.
Coordinate system and legend
All three plans use the same 480 × 360 cm room. Origin (0,0) is lower-left; x increases right and y increases up. All positions below are lower-left corners. Width is along x; depth is along y. The drawing scale is 0.9 image pixels per cm in both directions for every option. Screen y runs down, so an item at (x,y) with width w and depth d is drawn at (origin_x + 0.9x, origin_y - 0.9(y+d)) with size (0.9w,0.9d).
Furniture: D, desk, 140 × 70 cm; S, sofa, 180 × 85 cm; B, bookcase, 100 × 35 cm; T, side table, 50 × 50 cm. Blue/green/brown/purple fills identify these four items, and each item also carries its letter. K1 and K2 are dashed ochre keep-clear rectangles. Red hatched regions mark failed intersections or out-of-room strips. These encodings are repeated as text below; color is not the only key.
K1 spans x = 0 to 90 and y = 0 to 100 cm. K2 spans x = 360 to 480 and y = 260 to 360 cm. These user-supplied empty patches are arbitrary mock preferences, not real safety clearances. Edge contact is allowed; only positive-area intersections count as overlap.
Option A: satisfies the stated mock constraints
- Desk D: lower-left (120,260), upper-right (260,330).
- Sofa S: lower-left (240,40), upper-right (420,125).
- Bookcase B: lower-left (40,180), upper-right (140,215).
- Side table T: lower-left (170,140), upper-right (220,190).
- All four footprints are entirely within the room.
- All six furniture-pair intersection areas are 0 cm².
- All eight furniture-to-zone intersection areas are 0 cm².
A is the only supplied option that satisfies every stated mock rule. This does not establish comfort, usable circulation, accessibility or safety.
Option B: furniture overlap
- Desk D: lower-left (120,260), upper-right (260,330).
- Sofa S: lower-left (240,40), upper-right (420,125).
- Bookcase B: lower-left (40,180), upper-right (140,215).
- Side table T: lower-left (390,100), upper-right (440,150).
- All four footprints are within the room and both keep-clear patches are empty.
- Sofa S and side table T intersect at x = 390 to 420 and y = 100 to 125 cm.
- Intersection width 30 cm × depth 25 cm = 750 cm². The other five furniture-pair areas are 0 cm². All eight furniture-to-zone areas are 0 cm².
B fails the no-furniture-overlap rule. No item has been shifted to conceal the failure.
Option C: keep-clear intersection and outside boundary
- Desk D: lower-left (380,290), upper-right (520,360).
- Sofa S: lower-left (240,40), upper-right (420,125).
- Bookcase B: lower-left (40,180), upper-right (140,215).
- Side table T: lower-left (170,140), upper-right (220,190).
- Desk D reaches x = 520 cm, which is 40 cm beyond the room’s right boundary at x = 480.
- The out-of-room strip spans x = 480 to 520 and y = 290 to 360 cm: 40 × 70 = 2,800 cm².
- Desk D intersects K2 at x = 380 to 480 and y = 290 to 360 cm: 100 × 70 = 7,000 cm².
- All six furniture-pair areas are 0 cm². The other seven furniture-to-zone areas are 0 cm². The other three items are in bounds.
C fails two separate rules. Its furniture size is unchanged; the part beyond the boundary is deliberately shown.
Calculation method
For rectangles with x intervals [x1,x2] and [u1,u2], intersection width is max(0, min(x2,u2) − max(x1,u1)). Use the same rule for y intervals to find intersection depth, then multiply. A zero width or depth means zero overlap. A footprint is in bounds exactly when x ≥ 0, y ≥ 0, x + width ≤ 480 and y + depth ≤ 360. Preserve the source values before calculating. Do not infer room features from the sketch.
Perturbation: room width 410 cm, everything else fixed
This is a separate hypothetical change, not the base result. K2 still ends at x = 480 and would now extend 70 cm outside the revised room. The input contains an invalid keep-clear zone, so do not issue an overall pass or recommendation until the user resolves that changed source geometry.
The unchanged sofa ends at x = 420 in all three options, so it extends 10 cm beyond the new right boundary; its outside area is 10 × 85 = 850 cm². Option B’s side table ends at x = 440, giving a 30 × 50 = 1,500 cm² outside strip. Option C’s desk ends at x = 520, giving a 110 × 70 = 7,700 cm² outside strip. All other furniture is in bounds. The original furniture-pair intersections are unchanged, as are intersections with the unchanged zones. The revised room dimensions do not authorize moving the furniture or shrinking K2.
Before applying this to a real space
Check units, how dimensions were measured, footprint orientation and any measurement uncertainty. Ask for relevant omitted physical features and use requirements. The geometric result alone does not support claims about installation, structural work, accessibility, fire safety or comfortable use. The artifact review is limited to this synthetic coordinate fixture and its rendered image.
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.
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.
Treating a drawing’s top-left y position as the supplied lower-left room coordinate, which flips or shifts the footprint.
Resizing or rotating a furniture item, shrinking a reserved patch or clipping an out-of-bounds edge to make a picture look tidy.
Checking furniture against other furniture while forgetting the room boundary or personal keep-clear patches.
Calling boundary contact an overlap despite the explicit mock rule, or treating a positive-area intersection as harmless contact.
Adding apparently standard clearances that the source never supplied, or presenting a geometric fit as proof of safe or comfortable use.
Changing the source measurements while leaving the static CSV, image or text on the old values.
What has been checked?
Documentation
No dated source review recorded.
Included files
Scoped checks passed on 2026-10-14. Fixture checks; File parsing. See scope below.
AI workflow
Not run in a recorded consumer-AI workflow.
Time savings
Time savings not measured
Read the scope and evidence
layout-measurements.md
2026-10-14 · semantic fixture · passed
The fictional source specifies dimensions, units, axes, fixed orientation, every candidate, both arbitrary personal zones and the allowed edge-contact convention. No real-world layout or safety requirements are established.
480 × 360 cm room; x right and y up from lower-left
Four fixed furniture sizes and twelve candidate placements match independent controls
K1 and K2 match independently enumerated rectangles
Units, changed room, flipped axes, resized desk and shrunk zone mutations rejected
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
Plain UTF-8 CSV structure, stable header, row lengths and lossless parse/write/parse. No spreadsheet import or consumer tool tested.
One ten-field header and twelve ten-field data rows
Rows retain IDs, units and source order
No formula-triggering or control-character cells
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
Every coordinate, fixed size and computed edge is compared against independent fixture controls. Pair intersections, zone intersections, bounds and the 410 cm perturbation are recalculated separately.
A passes all stated mock rules; B overlap 750 cm²; C zone overlap 7,000 cm² and outside strip 2,800 cm²
Six furniture pairs and eight zone pairs per option checked
410 cm perturbation catches invalid K2 and outside furniture strips 850, 1,500 and 7,700 cm²
Five boundary-contact/one-cm tests pass; changed, missing and formula-bearing CSV mutations rejected
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
Self-contained XML parsed; bounded intrinsic dimensions and no active elements, external hrefs, embedded images or geometric transforms found.
1640 × 1020 viewBox and intrinsic dimensions
No scripts, foreignObject, external resources, events or transforms
No hidden content or additional unlisted rectangles
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
All 24 visible rectangles are independently recalculated from the source coordinates at 0.9 pixels per cm on both axes. Exact shape checks cover rooms, zones, furniture and highlighted failures; source review alone does not establish visual legibility.
Three rooms, six zones, twelve furniture rectangles and three derived failure rectangles match independent geometry
All x/y ticks, item label coordinates and conclusion labels match controls
Correct lower-left room to top-left SVG y conversion
Shifted, stretched, hidden, extra, transformed and relabeled visual mutations rejected
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
PNG signature, every chunk CRC, compressed raster payload and intrinsic dimensions checked. This parser check alone does not establish readability.
1640 × 1020 pixels
All PNG chunk checksums pass
Raster decompression succeeds and file ends at IEND
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
Actual final PNG directly inspected on 2026-10-14. All three plans, furniture IDs, zone labels, ticks, legend and exact failure summaries were legible without clipping. C’s full overrun and K2 label beneath the desk are visible. This is editorial image review, not a consumer AI workflow, browser, assistive-technology, real-room or safety test.
Actual final 1640 × 1020 PNG opened with view_image
Same-scale three-panel layout with C desk visibly beyond boundary
A clear, B overlap marked, C zone intersection and outside strip visible
All four furniture IDs and both zone labels per option legible; title, legend and scope note unclipped
Environment: Inkscape 1.4 SVG-to-PNG export followed by direct view_image inspection of the final raster. No image generation, consumer AI execution or browser test.
Text alternative and result matched against independently enumerated corners, recalculated intersections and the invalid-zone perturbation. Claims remain limited to the fictional geometry.
All twelve lower-left and upper-right corner pairs retained
A-only mock-constraint result; B 750; C 7,000 and 2,800 cm² claims correct
Changed room creates invalid K2; conditional outside areas 850, 1,500 and 7,700 cm² retained
Wrong corner, promoted option and incorrect perturbation mutations rejected
Environment: Local Python standard-library CSV/XML/PNG parsing and separately enumerated interval-geometry controls. Validator does not import generator data or functions. No consumer AI or browser test.
