# Pet doors & barriers: research notes

Date: 2026-10-25. Scope: two mechanisms bundled under one taxonomy leaf, kept as one register because the label bundles them, not because they are one thing. **Pet doors**: a flap or panel set into a door, wall, window or sliding-glass-door panel that lets a pet pass without a human opening anything, from a dumb flexible flap to a microchip-reading electronic door. **Barriers**: physical gates and panels that block an opening (doorway, stairway, archway, hallway) so no pet (and often no toddler) passes, pressure-mounted or screwed in. Split 1 (`product_type`) does the dividing the label does not.

Deliberately excluded, and left to siblings in `supplies > pets`: freestanding exercise pens and portable playpens that do not span a fixed opening (`pet-playpens`, Pet Playpens & Containment — overlaps hardest with "barrier"; the line I drew is whether the product spans and is fitted to a specific opening (gate) or stands free in open floor space (pen)); wireless/electronic correction barriers such as PetSafe's Pawz Away (collar receiver + radio transmitter, aversive correction) — these are behaviour-modification devices, not physical obstacles, and sit with `training-behavior` (Pet Training & Behavior, in_progress); crates, kennels and houses (`cages`, planned); carriers and travel crates (`pet-carriers`, planned). `catalog.json` confirms all four as siblings under `taxonomy_path: [supplies, pets, ...]`.

Quote labels: **verbatim** = I read the sentence myself in text I fetched with curl (HTML tag-stripped, or raw JSON); **fetched excerpt** = the WebFetch tool's model summarised or lightly reworded a page I did not read as raw text; **search excerpt** = from a search-engine summary. WebSearch reported its session budget exhausted (200 of 200) on the first query of this run, before any result was returned, so every source below was reached by direct fetch (curl or WebFetch on a known or guessed URL, or a URL found inside a sitemap). No Internet Archive snapshot was usable: `web.archive.org` returned no snapshot for the CPSC business-guidance page I wanted, and the WebFetch tool refused to fetch `web.archive.org` entirely. **Could not reach**: cpsc.gov (returns Access Denied / 403 to curl and 404 to WebFetch on every URL I tried, including the business-guidance page on expansion gates); astm.org (403, and the standard text is paywalled regardless); petsafe.net/petsafe.com's own `/pages/how-to-measure-your-pet-for-a-pet-door` and `idealpet.com`, `halepetdoor.com/pages/sizing-guide`, `plexidor.com`, `dogdoors.com`, `petdoors.com`, `carlsonpetproducts.com/pages/how-to-choose-the-right-pet-gate`, `northstatesind.com/pages/pet-gates-buying-guide`, `richellusa.com/pages/pet-gate-guide` — all guessed slugs that 404'd; I did not have search to find the real ones. Working URLs for petsafe.com and carlsonpetproducts.com were recovered from each site's own sitemap.xml.

## Seed

Example given: car-tyres `treadwear_grade` — a number on every listing that still needs a judgement, because it only compares within one manufacturer's range. Persona: an audio engineer, who thinks in measurable versus perceived quality and where marketing lives in the gap.

1. **The treadwear-grade equivalent here is a pet door's "step-over" rule and a barrier's height.** Both are numbers a brand will give you (flap size by weight; gate height up to 41 in) that look like simple specs but are really a judgement about one specific animal: a flap sized right for a 15 lb cat is wrong for a 15 lb terrier with a deep chest, and a gate height that stops one dog is a launch pad for another. The number is comparable within a brand's own chart (Endura Flap's five sizes scale together) but not across brands, because no two brands measure "shoulder height" the same way and no standard body publishes a reference curve the way UTQG does for tyres. **Kept**: the register treats pet size (weight plus, for doors, shoulder height) as a split, and states plainly that flap/gate size numbers are brand-specific, not interchangeable. **Dropped**: I looked for a numeric rule tying gate height to a dog's measured jump height and found none sourced (only Carlson's gate-selector UI, which lets a buyer type a height up to 41 in but asks nothing about the dog) — no formula goes in the JSON.
2. **The audio engineer's instinct — where does the marketing number depart from a measured one? — lands on two single-brand efficiency claims.** PetSafe's Extreme Weather Door is "Rated 3.5 times more efficient than single flap doors" (verbatim, their own product page) and Endura Flap claims a "50+ mph wind rating" and flexibility "to −40°F" (fetched excerpt, their own site). Both are real, specific-sounding numbers with no visible test method and no independent source — exactly the gap the persona is tuned to notice. **Kept**: both appear only as one-brand claims with the number removed from any facet range and placed instead in `why_experts_care`/notes as "manufacturers claim," never as a backed threshold. **Dropped**: a prospective `insulation_r_value` facet — no brand publishes an R-value for a pet door, so there is nothing to put a number on.

## Sources and what each supports

### Regulator

**16 CFR Part 1239, Safety Standard for Gates and Enclosures (law.cornell.edu e-CFR, read as raw HTML text)**
https://www.law.cornell.edu/cfr/text/16/part-1239 ; https://www.law.cornell.edu/cfr/text/16/1239.1 ; https://www.law.cornell.edu/cfr/text/16/1239.2
- verbatim §1239.1: "This part establishes a consumer product safety standard for gates and enclosures."
- verbatim §1239.2: "Each gate and enclosure must comply with all applicable provisions of ASTM F1004-25, Standard Consumer Safety Specification for Expansion Gates and Expandable Enclosures, approved on December 1, 2025." Source note: "85 FR 40112, July 6, 2020."
- Supports: that a CPSC-mandated US standard exists for gates and enclosures and names ASTM F1004-25 as the incorporated text. Confidence high for the citation itself. I could not read ASTM F1004's actual clauses (paywalled at astm.org, which also 403'd to curl) or CPSC's own plain-language summary (cpsc.gov blocked every URL I tried), so the register does not state any specific gap, height or force number from the standard — only that it exists and that no pet-gate brand page I reached names it. This is one source for the existence of the standard; I did not find a second independent source confirming ASTM F1004's content because I could not read the content at all, so the "governing standard" facet states only what the regulation itself says (scope, incorporation, title/date) and nothing about what it requires.
- Judgement: the standard's scope line names "gates and enclosures" without a child/pet carve-out, but every durable-infant-product part in the same CFR subchapter (1215–1243) is about child products, and none of the pet-gate product pages I reached (Carlson's collection page, its gate-selector quiz) mentions ASTM F1004 or CPSC compliance. Low-to-medium confidence judgement: the mandatory standard binds gates marketed for child safety; pet-only gates are a grey area, often similar hardware marketed without the claim.

### Brand spec sheets — pet doors

**Sure Petcare, agent product API (JSON, read as raw text; same host as the public site, so I treat it as that brand's own published specification, like a spec-sheet page)**
https://www.surepetcare.com/api/agent/v1/products/3 (Microchip Cat Flap), /4 (DualScan Microchip Cat Flap), /97 (Microchip Pet Door Connect)
- verbatim (JSON field values, Microchip Cat Flap, product 3): `"Door Opening Dimensions W(mm) x H(mm)": "142 x 120"`, note "This dimension is the smallest area of the door that your pet will have to enter"; `"Installation Hole Dimensions W(mm) x H(mm)": "165 x 171"`; `"Number of pets": "32"`; `"Batteries Required": "4 x AA"`, note "Typical battery life is over six months, and in many cases, up to 12 months with normal use."; `"Manual Override Lock": true`; `"Electronic Override Lock": false`; `"Selective Entry": true`, note "Will only allow registered animals to enter your home"; `"Selective Exit": false`; `"Integrated Draft Excluder": true`.
- verbatim, DualScan Microchip Cat Flap (product 4): same opening dims (142×120 mm), but `"Selective Exit": true` and still `"Electronic Override Lock": false`, `"Remotely lock/unlock the door": false`.
- verbatim, Microchip Pet Door Connect (product 97, the dog-size version): `"Door Opening Dimensions W(mm) x H(mm)": "178 x 170"`; `"Installation Hole Dimensions W(mm) x H(mm)": "207x234"`; `"Electronic Override Lock": true`, note "Controlled using the built in LCD, you can set the door to stop pets leaving or entering the house at any time."; `"Remotely lock/unlock the door": true`; `"Entry/exit notifications": true`; `"Timer Control": true`, note "The pet door can be programmed to lock and unlock at specified times of the day."; `"Batteries Required": "4 x C"`.
- verbatim, compatibility note on product 3: "suitable_pets": "Cats only (flap opening 142 x 120 mm)", reason "Confirmed by the product owner: the cat flaps are sized for cats and are likely to be too small for even a very small dog... If the pet is a dog, the Microchip Pet Door or Microchip Pet Door Connect is the one to recommend instead - it opens 178 x 170 mm."
- Supports: the three-tier locking family (manual override only / manual + selective entry+exit / full electronic with app-style LCD, remote lock, notifications, timer) that the register's `access_control` facet is built on; the opening-dimension-as-ceiling framing of sizing; battery type and life for electronic doors; that a single identity (chip or collar tag) works across every Sure Petcare product in a house. One brand, high confidence for what it states about its own products; no second brand publishes the same fields in the same form, so cross-brand comparison stays qualitative in the register.

**PetSafe, product pages (petsafe.com; real URLs recovered from `sitemap-0.xml`, since guessed `/pages/...` slugs 404'd)**
https://www.petsafe.com/p/4-way-locking-cat-door/P1-4W-11/ ; https://www.petsafe.com/p/electronic-pet-door/PPA11-10709/ ; https://www.petsafe.com/p/extreme-weather-doors/HPA11-10987/ ; https://www.petsafe.com/blog/is-your-pet-door-secure/
- fetched excerpt, 4-Way Locking Cat Door: the four lock settings are "locked, unlocked, in-only and out-only"; flap opening 5½ × 6⅛ in, frame 7½ × 8 in, cutout 6½ × 7 in; fits "doors 1 1/8 inches - 2 inches thick."
- fetched excerpt, Electronic Pet Door: identification is "1 Electronic Pet Door Key that attaches to your pet's collar," up to "five programmable SmartKeys"; door "uses 4 D-cell batteries, not included"; key "uses 1 PetSafe RFA-67 battery, included"; two sizes, small "5 1/2 x 7 7/8 opening for pets up to 15 pounds," large "11 x 16 opening for pets up to 100 pounds."
- verbatim (confirmed by curl on the raw page, not just the WebFetch summary), Extreme Weather Pet Door: "Rated 3.5 times more efficient than single flap doors, this door blocks severe temperatures with 3 flaps, incl[uding a center insulated flap for added protection]." Sizes (fetched excerpt): small 5⅛ × 8¼ in for pets up to 15 lb, medium 8⅛ × 12¼ in up to 40 lb, large 10⅛ × 16¼ in up to 100 lb. No numeric temperature range is given anywhere on the page.
- fetched excerpt, "Is Your Pet Door Secure?" blog post: "it is possible for a human to fit through some large pet doors"; advice is to "buy the smallest one your dog or cat will fit through," to prefer wall installation over a door-mounted unit ("Wall-installed doors are more secure... and are harder to crawl through"), to use a collar-sensor locking door, and to "lock your pet door securely during the day since that is when most burglars are active."
- Supports: a second, independent locking-mechanism family — collar-worn RF key (distinct from Sure Petcare's chip/tag reading) — with its own battery facts; the "4-way lock" vocabulary term, which matches the existing add-on schema's `four_way_lock` enum value exactly; pet-weight-based sizing from a second brand (small ≈ up to 15 lb, medium/large scaling to 100 lb), independently confirming the general fact that door size is sold by pet weight even though the brands' own numbers don't line up with each other or with Endura Flap's; the security judgement (single brand, its own blog, no independent test — kept as a medium-confidence judgement, not a fact). The "3.5 times more efficient" and the temperature claims are single-source marketing numbers with no stated method; they do not appear as a sourced range or threshold in the JSON, only as a quoted claim in prose.

**Endura Flap, product pages (enduraflap.com; curl raw HTML, WebFetch summaries checked against the raw text for the size chart)**
https://www.enduraflap.com ; https://www.enduraflap.com/pages/measure-your-pet ; https://www.enduraflap.com/pages/find-an-endura-flap-pet-door-that-fits-any-pet
- verbatim (confirmed in raw HTML, matches the WebFetch summary exactly): size table — "Flap size (W × H)": Small 6″×11″, Medium 8″×15″, Large 10″×19″, XL 12″×23″, XXL 14″×27″; "Max shoulder height": up to 18″/22″/27″/31″/37″; "Minimum weight": 10/20/20/30/50 lb.
- verbatim, measuring instructions: "measure from the floor to the tallest point of your dog's back (this may be their shoulders, mid-back, or tailbone)"; step-over rule — "The top of the flap should be mounted at least 1" higher than this [combined height]... In no case should the step-over exceed 1/3 of the height at the withers – the lower the step-over, the better."
- fetched excerpt, home page: "50+ mph wind rating"; flexible material "to −40°F"; magnetic sealing to reduce drafts; double-flap models add "an additional layer of insulation"; frames in white aluminum or bronze; installs in wood, metal, fiberglass and storm doors.
- Supports: a third brand's own weight-and-height sizing chart (so the general "door size scales with pet weight and height, chart is brand-specific" fact now has three independent brands behind it: Sure Petcare, PetSafe, Endura Flap); the step-over/withers rule, kept in the register as a single-brand rule of thumb (no second source found, so it is a medium-confidence judgement, not a derived-metric formula); the wind/cold ratings, kept as single-brand marketing claims only, same treatment as PetSafe's efficiency number.

### Brand pages — barriers (pet gates)

**Carlson Pet Products (carlsonpetproducts.com; curl plus WebFetch, canonical host without `www.`)**
https://carlsonpetproducts.com/pages/gate-selector ; https://carlsonpetproducts.com/collections/gates
- fetched excerpt, gate-selector quiz: four questions — "WHERE DO YOU NEED A GATE?" (pressure-mounted for a doorway/stairway, or freestanding for an open space); "WHAT MATERIAL(S) DO YOU WANT YOUR GATE TO BE?" (metal, wood or plastic); "HOW WIDE IS THE SPACE...?" (23 to 192 in); "DOES YOUR GATE NEED TO HAVE A MINIMUM HEIGHT?" (up to 41 in). The quiz does not ask about the dog at all — no weight, no jump history.
- fetched excerpt, gates collection: named products and their mount type — "Extra Wide Walk-Thru Pet Gate" and "51 Extra Wide Walk-Thru Pet Gate" (pressure-mount, steel, with a small walk-through pet door and, on the 51 in model, an optional hardware-mount); "Plastic Expandable Pet Gate" (23.5 in tall, fits 28–42 in, pressure-mount, plastic with rubber bumpers); "Convertible Pet Yard" (28 in tall, expands 52–146.5 in, steel, slide-handle walk-through door). No ASTM or other safety-standard mention found on the collection page.
- Supports: that pressure-mount versus freestanding (not pressure versus hardware-mount, which is the framing child-gate guidance uses) is the live choice a pet-gate seller puts in front of a buyer; that width range and height are both sold as hard numbers on the listing; that a walk-through door is a named, filterable feature; that chew-resistance is sold as a material fact (steel marketed as "chew-resistant," plastic is not). One source; I could not reach a second pet-gate-specific brand's buying guide (Richell, North States and PlexiDor's equivalent pages all 404'd on guessed URLs), so the register treats "ask about mount type, width, height, material" as this brand's own framing, corroborated only by the general, widely-known pressure/hardware distinction rather than a second brand's explicit guide.

**PetSafe, wireless indoor barrier (kept out of scope, recorded for the boundary call above)**
https://www.petsafe.com/blog/pawz-away-indoor-barrier-to-the-rescue/
- fetched excerpt: a transmitter creates an adjustable-range zone; the pet wears a collar that gives "a warning sound" then "it feels like someone pushing you away" if the pet crosses; used at windows, trees, furniture. Confirms this product is a correction device (collar + radio transmitter), not a physical obstacle, which is why it is scoped to `training-behavior`, not here.

## Data points used for derived metrics

No `derived_metrics` entries. I looked for a number experts compute in their head for this category (the tyres/power-bank style "cost per X") and did not find one: price is not computed per inch of opening or per pound of pet by anyone I read, and the one candidate formula (step-over ≤ ⅓ of shoulder height) has one source, so it is a judgement, not a formula with `depends_on` facets. A thin derived-metrics section is correct here, not a gap to fill.

## Against the existing add-on schema

Add-on read: `supplies/pets/add-ons/pet_doors_barriers.jsonc` — `locking_mechanism` (enum: manual_slide, magnetic, electronic, microchip_rfid, four_way_lock, none, unspecified), `vehicle_fit` (string), `sizing.flap_size` (string, "the fit gate"). Composed with `pets-standard.jsonc` (`animal_type`, `sizing.suits_weight_range`, `sizing.max_pet_weight`) and `supplies-standard.jsonc` (`high_signal_markdown`, `sizing.dimensions`, `sizing.pack_count`, etc). Field-kind vocabulary from `zamana-expertise/AGENTS.md`: vertical anchor, category quality currency, hard filter, sizing, or cut to `high_signal_markdown`.

| Facet in this register | In add-on today | Proposed kind (AGENTS.md) | Note |
|---|---|---|---|
| `product_type` | no (implicit in category) | — | Not a product field; it is the register's own top split, since the leaf bundles two product kinds. Nothing to add to the schema. |
| `pet_size_category` | partly — `pets-standard.sizing.suits_weight_range` / `max_pet_weight` already exist | sizing (already correct) | The register adds the judgement (brand-specific charts, no cross-brand standard), not a new field. No change proposed. |
| `access_control` (door locking tier: any animal / manual lock only / selective entry) | partly — `locking_mechanism` enum covers the categorical values reasonably well (`magnetic`≈any animal, `four_way_lock`≈manual lock with in/out, `microchip_rfid`≈selective entry) | hard filter (already there) | Keep the enum. |
| `remote_features` (app/LCD control: remote lock, curfew timer, entry/exit alerts, beyond basic selective entry) | no | high_signal_markdown | Only Sure Petcare's Pet Door Connect and PetSafe's app-linked doors have this; rare, brand-spec-only, so it is a markdown line, not a column. |
| `installation_surface` (exterior door / wall / sliding glass / interior door / window) | no; closest is `vehicle_fit` (irrelevant to this reading — likely a leftover enum name; no pet door in any source here mentions a vehicle) | hard filter | Proposal: add `installation_surface` enum. Titles usually state it ("for sliding glass doors"), so listing with low backfill cost; `vehicle_fit` looks like noise for this category (see below). |
| `flap_opening_dimensions` (W×H) | yes, as `sizing.flap_size` (string) | sizing (already correct) | Keep. Register's data (Sure Petcare, PetSafe, Endura Flap) shows this is always a paired W×H, so a structured `{width, height, unit}` would parse cleaner than a free string, but that is a format nit, not a new field. |
| `flap_construction` (single flexible flap / double flap / rigid insulated multi-flap panel) | no | high_signal_markdown | Brand marketing copy states this reliably ("3 flaps," "double-flap"), titles sometimes do; researched, medium cost. Drives the (single-source, unverified) efficiency and cold-tolerance claims, so keep it a markdown line, not a currency number. |
| `pet_count_supported` (max chips/keys the unit stores) | no | annotation, researched | Only relevant to multi-pet homes; brand spec pages only (Sure Petcare: 32; PetSafe: 5 keys). Low priority. |
| `battery_requirement` (none / batteries, with type) | no | high_signal_markdown | Reliable on brand pages, rare on listings; researched, low cost once on the brand page. |
| `gate_height_in` | no | hard filter | Proposal: add. Every gate brand prints height; it is the primary filter against a specific dog's jumping/climbing. Listing, low backfill cost. |
| `gate_opening_width_range_in` | partly — `supplies-standard.sizing.dimensions` is a single length/width/height/depth block, not a fit range | hard filter / sizing | Proposal: a `{min, max, unit}` fit range belongs in `sizing`, since it is the bought-by-number dimension (AGENTS.md rule 2 and the sizing dimension test), not a single L/W/H triple. |
| `gate_mount_type` (pressure-mounted / hardware-mounted / freestanding panel) | no | hard filter | Proposal: add. Reliably on the listing (title or bullet), low cost. |
| `gate_material` (steel / wood / plastic / mesh-fabric) | no | hard filter | Proposal: add. On the listing; drives the chew-resistance judgement. |
| `walk_through_door` | no | hard filter (boolean) | Proposal: add. A named feature on listings ("walk-through gate"), low cost, and not a photo fact a buyer would reliably eyeball from a thumbnail. |
| `meets_gate_safety_standard` | no, and the add-on has no certification field at all | high_signal_markdown | Rare on listings for pet-only gates (no source here shows a pet-gate brand citing ASTM F1004); keep as a markdown line when a listing does state it, per AGENTS.md's general rule that certifications are cut to markdown, not a column. |

Add-on fields this register considers noise for this category: `vehicle_fit` — no source read for this register (pet door or barrier) mentions a vehicle; it looks like a field copied in from another pet add-on by mistake, or scoped to a product type (pet car-door inserts / cargo-area barriers) that none of the brands I reached sell as a named line. I would not cut it without checking whether some listings in the actual catalog use it, but nothing in this research supports it.

## Branches

Traced to a specific sentence in the register or these notes; any leg I could not trace this way was dropped.

- `product_type = door` → `then_ask: [access_control, installation_surface, flap_construction]`, `makes_irrelevant: [gate_height_in, gate_opening_width_range_in, gate_mount_type, gate_material, walk_through_door, meets_gate_safety_standard]`. Traced to the scope paragraph above: a door and a barrier are different products built for opposite purposes, so a barrier's mount type, height, width-fit and gate standard have no value once the buyer has said they want something their pet can use alone. (`meets_gate_safety_standard` was added to the register after this trace was first written; it follows the same door/barrier line because the standard it names is titled "gates and enclosures," not doors.)
- `product_type = barrier` → `then_ask: [gate_mount_type, gate_height_in, gate_opening_width_range_in]`, `makes_irrelevant: [access_control, installation_surface, flap_construction, pet_count_supported, battery_requirement, remote_features]`. Same trace in reverse: a gate has no locking tier, no installation surface in the door sense, and (battery-free, in every source read here) no battery or remote-control question. (`remote_features` added for the same reason as above.)
- `gate_mount_type = pressure_mounted` → `then_ask: [gate_height_in]` with a note (not a hard rule, since I have no CPSC text to cite) that a pressure-mounted gate can be pushed over, which matters more the taller and heavier the dog. Traced to Carlson's own quiz framing it as pressure-mount for "doorways/stairs" versus freestanding for open space, plus the 16 CFR 1239 / ASTM F1004 citation that *some* gate standard exists for exactly this failure mode in the child-safety version of the product — I did not find the clause, so the branch note stays a flagged caution, not a sourced rule, and I did not add a `makes_irrelevant` leg here because I have nothing specific it removes.
- `access_control = any_animal` (a dumb flap with no locking) → `then_ask: []`, with the security judgement from PetSafe's own blog post surfaced as a note on `installation_surface` (wall beats door-mounted) and on `pet_size_category` (buy the smallest size the pet needs). Traced directly to the quoted PetSafe sentences above. I did not add a `makes_irrelevant` leg, since nothing stops mattering, only something gets added (the security note).

Dropped leg: a "pet count supported" branch off `access_control = selective entry`, asking how many pets the household has. I considered it because Sure Petcare's 32-chip and PetSafe's 5-key limits only matter once selective entry is chosen, but a multi-pet household is just as relevant to a `none`/dumb-flap door (where the facet is irrelevant for a different reason: a dumb flap has no limit at all) as to a selective one, and the branch would not change what else gets asked, only add one more fact — so it stays a plain annotation facet with no branch, per "only write a branch where the answer really changes the path."

## Trimmed detail

- PetSafe's full "Is Your Pet Door Secure?" post also recommends generic home-security measures (alarm systems, motion-sensor lighting, cameras) unrelated to the product category; left out of the register as the generic-test would catch it (it would read the same for a window lock).
- Endura Flap's frame-material and installation-surface compatibility list (wood, metal, fiberglass, storm doors) is folded into the `installation_surface` facet's definition rather than given its own facet; it is a feature list, not a split.
- The Sure Petcare API's accessory/compatibility data (tunnel extenders, mounting adaptors, glass-mount hole sizes 212–260 mm vs 294–320 mm) is real and sourced but is installation-accessory detail, not a buyer-facing facet; omitted from the JSON, kept here in case a later pass wants an `accessories` facet.

## Corrections

- `validate.py` rejected the `access_control = any_animal` branch ("changes nothing. A branch must name at least one facet to ask next or to drop"), because the security caution that leg carries is a note on two other facets, not a `then_ask`/`makes_irrelevant` change. Removed the branch; the security judgement stays attached to `installation_surface` and `pet_size_category` as written. No other validator, render or viewer errors. (First pass — the `pet-doors-barriers/` output folder was empty before this run, so there is nothing else to correct against.)

## Corrections (2026-10-26)

- Fetched five Carlson gate product pages (Big Tuffy, Tuffy, Plastic Expandable, Extra Wide Walk-Thru, 51" Extra Wide Walk-Thru; curl into `scratch/fix-pet-doors-barriers/`) and confirmed verbatim each states "...has been certified by the American Society for Testing and Materials (ASTM)," with no standard named; Lil Tuffy, Convertible Pet Yard and Direct View carry no such line. Rewrote judgement 2's claim/note from a flat negative to this inconsistency and added `carlson_tuffy_gate`/`carlson_big_tuffy_gate` as new references.
- Rewrote `meets_gate_safety_standard`'s `why_experts_care` to state the Carlson ASTM-without-standard-name pattern instead of "no pet-only gate brand cites it."
- Fixed `meets_gate_safety_standard`'s `backfill_cost` (high → low) and `listing_field_hint` (the claim sits in the ordinary product description, not a separate compliance page).
- Added a `backfill_note` to `gate_mount_type` stating Carlson's "Expandable" line is pressure-mounted in every case read, so it folds into `pressure_mounted` rather than needing a fourth value; same note flags all three values as single-brand (Carlson only).
- Folded the dropped `gate_mount_type = pressure_mounted → then_ask: [gate_height_in]` branch leg (re-asked a facet already asked on every barrier path) into `gate_mount_type`'s `why_experts_care` as a one-clause note that push-over risk scales with the dog's height and weight; removed the branch.
- Added a Sure Petcare Hub-dependency clause to `remote_features`'s `why_experts_care` (remote lock/unlock and entry/exit alerts need the separate Hub; Timer Control and Electronic Override Lock do not) and raised its `backfill_cost` medium → high with a note explaining why.
- Added a tie-breaker sentence to `product_type`'s definition for Carlson's Convertible Pet Yard (gate here when sold to span a fixed opening, `pet-playpens` when sold to stand free), resolving the unreconciled sibling-scope overlap.
- Rephrased `pet_size_category`'s cold-start question (and the matching `splits` entry) from "Small, medium, large, or giant breed?" to "About how much does your pet weigh, in pounds?" — the old question was generic (would read the same on a collar or crate register) and didn't map to any stated weight/height cut points; weight ties directly into the facet's own `source: derived` bucket logic.
- Added single-source `backfill_note`s to `gate_height_in`, `gate_opening_width_range_in`, `gate_material`, `walk_through_door` and `pet_count_supported`, flagging the 41 in / 23-192 in Carlson bounds, the Carlson-only barrier facet values, and the 32-pet/5-key limits as single-brand figures not cross-checked against a second brand.
- Rejected check.md's Part D note on "Source: listing" for eight facets being unverified against a third-party retail listing: this mirrors a known cross-category limitation (retail listings block fetches for every author), not a defect specific to this register's sourcing, and brand-page content is the stated basis already; no change made.