Diagnose first. Sequence second. Do not invent the missing number.

Lab uses 36 inches as a conservative planning fail line for a primary walking route and 42 inches as preferred breathing room. The 36-inch benchmark is consistent with the ADA continuous accessible-route minimum and the IRC hallway minimum, but Room Lab does not decide whether ADA, IRC, or another local rule applies to a particular dwelling.

2010 ADA Standards §403.5.1 (36 in continuous clear width on an accessible route), per the U.S. Access Board guide; 2024 IRC R318.6 hallway width, 3 ft (R311.6 in the 2021 IRC) · source · checked 2026-09-24 · review by 2027-09-24 · vlab.clearance.v2

Both cited figures are code minimums for specific contexts — an accessible route in a facility the ADA covers, and a hallway in a dwelling — not residential comfort rules for the path between furniture. Lab borrows the number as a planning benchmark. Door clear width, accessible turning geometry, required egress, and local amendments are separate questions and are never certified here.

• Whether a given route is legally required egress or an accessible route depends on the building, occupancy, jurisdiction, and scope of work.

Lab uses 42 inches as the planning threshold for a one-cook work aisle. The cited NKBA guideline recommends at least 42 inches for one cook and at least 48 inches for multiple cooks, measured between counter frontage, tall cabinets, and appliances. Room Lab does not convert that recommendation into a building-code claim.

NKBA Kitchen Planning Guidelines with Access Standards — Guideline 6: Work Aisle (earlier, freely published edition) · source · checked 2026-09-24 · review by 2027-09-24 · vlab.clearance.kitchen.v2

This is the most recent edition of the figure that can be checked publicly. NKBA has since published later editions, including the 5th edition in 2026, whose full text is sold rather than published; they reorganise circulation clearances, so the current figure may differ. Measure between the occupied frontages that define the aisle. Appliance doors, handles, and people can change what remains usable in practice.

• Whether NKBA's 5th edition (2026) keeps 42 / 48 inches for the work aisle. Its public summary of changes says circulation clearances were consolidated into a new Guideline 1.5 with errata corrections and code updates, but does not state the figures.

Lab uses 14 inches as a compact reach-and-knee-room benchmark when evaluating a coffee-table relationship. It is a comfort heuristic, not an accessibility or code threshold; users should test the actual seat depth, body fit, and table use.

Crate & Barrel coffee-table buying guide (keep the table about 14 inches from the sofa); Houzz living-room measurements (as close as 12 inches, within 18–20 inches to reach) as a cross-check · source · checked 2026-09-24 · review by 2028-09-24 · vlab.clearance.seat-table.v2

Retailer and magazine guidance, not a standard. Published furniture guidance commonly places this relationship in roughly the 12–18 inch range.

Lab uses 36 inches as a minimum planning band behind dining seating. It is enough to test chair pull-out and basic movement, but it is not a promise that a person can circulate comfortably behind an occupied chair in every configuration.

West Elm Rules of Thumb: Dining Room (36–44 inches from table to wall to pull out chairs; 54–62 inches with a buffet) · source · checked 2026-09-24 · review by 2028-09-24 · vlab.clearance.dining.v2

More clearance may be necessary when people must pass behind occupied chairs, when a buffet or cabinet also needs operating space, or when accessibility requirements apply.

Lab uses 30 inches as the planning floor for the space in front of a door, measured from the door wall to the nearest declared piece, the opposite wall, or the edge of a non-rectangular floor — whichever is closest. Where that distance cannot be measured the finding is withheld rather than assumed.

Room for Drama planning heuristic — the floor a person needs in front of a door to open it and stand there. Not taken from a code landing requirement. · checked 2026-09-07 · review by 2028-09-07 · vlab.clearance.door-landing.v1

Required landings are a different and larger question. Egress landings, accessible manoeuvring clearances at doors, and the swing of the leaf itself are each governed by their own criteria, vary with approach direction and jurisdiction, and are never certified here.

• Whether a given door is required egress, or sits on an accessible route, is not decided by Room Lab — and both carry clearances larger than this heuristic.
• The arc the leaf sweeps is modelled separately, and only when the record says which side the hinge is on.

Lab uses 24 inches as a compact planning threshold beside a bed. Current designer guidance varies with bed size, number of sleepers, furniture, mobility, and whether the side is a true circulation route; more space is usually more comfortable.

Homes & Gardens, expert guide to bedroom clearances (designers quoted: 18–24 inches each side for a twin/full, 24–30 for a queen, 30–36 for a king) · source · checked 2026-09-24 · review by 2028-09-24 · vlab.clearance.bed.v2

This is intentionally not a universal minimum. A required accessible route or maneuvering space is governed by different criteria.

Two questions, two figures. 72 inches is whether a person can stand under the low side. 78 inches is whether a full-height piece — a wardrobe, a standing desk, a tall shelf — can go against it. A wall can pass the first and fail the second, and Lab now says so in those terms rather than calling both of them "standing".

Room for Drama planning heuristic, benchmarked against the 2024 IRC R305 habitable-room ceiling height (7 ft, with allowances for sloped ceilings) and against published wardrobe and shelving heights · checked 2026-09-24 · review by 2028-09-24 · vlab.headroom.v1

Neither figure is the code minimum for a habitable room, which is higher and is measured over a proportion of the floor area rather than at one wall. Whether a room with a sloped ceiling counts as habitable, and how much of its floor area counts, is a question for the jurisdiction.

• Room Lab does not determine habitability, required ceiling height, or how much of a sloped room's floor area counts toward it.

Lab treats about eight feet as a conversational-group benchmark, not a hard boundary. Larger rooms may work better as multiple seating groups; acoustics, hearing, furniture orientation, and the actual people in the room can change the useful distance.

Houzz, key measurements for your living room (eight feet is ideal; larger rooms may need several seating groups) · source · checked 2026-09-24 · review by 2028-09-24 · vlab.conversation.v2

This threshold influences composition guidance only; it does not gate safety or code compliance.

The full reading scores eight dimensions, and pressure on a dimension is higher-is-worse. The bench ranks six planning-condition pressures the same direction, high-to-low, and tie-breaks movement → function → light → comfort → reset → personality. The front door's six-question Room score is a different number and runs the other way: higher is better. A missing score does not rank and is not treated as zero.

The reading engine (RFD-RP-001), projected onto the six planning conditions · source · checked 2026-09-07 · review by 2027-09-07 · vdiagnosis wire v1 (file format pulse.diagnosis.v1, kept for import) / envelope 1.1.0

Legacy 0/1/2 ratings map to pressure 100/50/0. Quick Start does not invent scores.

Structural, electrical, gas, moisture, and the way out are outside Lab authority when they fail. A watch is visible and does not lock. Unknown is unknown — Lab does not invent a fail or an all-clear.

Room for Drama methodology — Lab will not sequence atmosphere over a named safety concern · checked 2026-09-07 · review by 2027-09-07 · vlab.gate.v1

The professional-boundary status is mapped so a named safety concern cannot be skipped.

Where these zones sit is measured — a drop zone is anchored to a declared door, a buffer to the edge between two placed pieces. How deep they are drawn is a convention, because neither is an object the room can be asked about: two feet is a landing rather than a store, one foot is a curtain or a shelf edge rather than a wall. Every other zone depth is taken from CLEARANCE and published above.

Room for Drama spatial method — zone conventions, not clearances · checked 2026-09-07 · review by 2028-09-07 · vlab.zones.v1

These set the size of a rectangle on a drawing, never a pass or a fail. No clearance verdict is computed from either figure.

Architecture is given. Furniture, light, planting, pattern, and soft objects stack on it. A layer that does not reduce the primary friction is held or refused.

Room for Drama spatial method · source · checked 2026-09-07 · review by 2027-09-07 · vlab.layers.v1

Atmosphere biases the upper stack; it never unlocks architecture.

A reversible test is information. Two simultaneous changes make drift unreadable. Ask the same questions again after seven days. If the primary did not move, put the change back.

Room Lab 7-day test · checked 2026-09-07 · review by 2027-09-07 · vlab.experiment.v1

Duration is a test window, not a clinical finding.

The reading is filed in this browser and never leaves the tab during the reading-to-bench handoff. An imported reading is shown to you on a plate before it loads. Unknowns stay unknowns.

Shared rules for what travels · checked 2026-09-07 · review by 2027-03-07 · v1

applicationId RFD-RP-001. Browser-local is the default. A separate, explicit Room Record control can optionally keep a server copy for cross-device use.

Room Lab is fully usable with sync off. If a person explicitly turns on a server copy, the browser creates the recovery key; the service stores a one-way SHA-256 hash of that key and the record payload. There is no Room Lab account or email recovery path, so losing the key loses access to that server copy.

Room Lab Room Record service — Supabase Edge Function + Postgres, keyed without an account · checked 2026-09-07 · review by 2027-03-07 · vroom-record.v2

Direct table access stays closed behind row-level security; the Edge Function is the application boundary. Turning sync off on one browser is different from deleting the server copy for every device.

Width, depth, height, material, price, lead time, and care are captured from the publisher or entered by the occupant. A missing number stays unknown. Catalog seed is IKEA US pages captured 2026-09-21. Fit tests scale, circulation (36" heuristic), visual mass, declared jobs, and whether the layer reduces the primary friction. A picture is not a measurement; Concept View is later and labelled concept.

Room for Drama methodology — a product is tested against this room's recorded dimensions, never against a mood · source · checked 2026-09-07 · review by 2027-09-07 · vlab.spec.v1

URL fetch is SSRF-safe https only. Published product pages are preferred. IKEA measurement lines are a fallback. Prices are a snapshot, not a live quote.

• Lead time is ZIP- and stock-dependent and is never invented as a single number.
• Whether a given door is required egress remains unknown unless the reading marked it as needing professional review.

A named budget is architecture. Known commitments are taken first. A reserve is held as unknown-cost, not leftover paint. Remaining money is sequenced free → test → commit → lock. The ranking, whether professional review is required, and measured-room area score four profiles (Balanced, Atmosphere-forward, Function-first, Lean). Atmosphere-forward waits while movement is the primary problem or a named safety concern is on. An unnamed budget is unknown, not zero. Spec Desk published prices are tested against remaining money; they are snapshots, not quotes. Thresholds are heuristics.

Room for Drama methodology, theory shared with Tangled Thistle Budget-to-Atmosphere, rewritten for rooms · source · checked 2026-09-07 · review by 2027-09-07 · vlab.budget.v1 / lab.budget.intel.v1

Event categories from Budget-to-Atmosphere (venue, food, guests) are not used. This is a planning model, not a specification and not a code opinion.

• Landed cost (freight, tax, haul-away) is asked as a hidden-cost prompt and is never invented as a number.
• Per-square-foot bands are planning heuristics, not market comps.