# Rolling pins: research notes

Date: 2026-10-25. Scope: every hand tool sold for rolling dough flat: handled (American or baker's) rollers, handleless straight dowels and tapered French pins, short thin pins for dumplings, roti and tortillas (Chinese pin, belan), long pasta pins (mattarello) and the oklava, plastic fondant rollers, textured, ridged and embossed pins (lefse, tutove, patterned cookie pins), pins with thickness rings or built-in guides, children's pins, and rolling-pin spacer rings sold alone. Left to siblings: pastry mats and silicone rolling mats (`baking-sheets`, "Mats & Liners"); pastry boards, chakla and bench surfaces (`kitchen-boards`, "Prep Surfaces"); cookie cutters, fondant cutters, smoothers and stencils (`molds`, "Cutters & Stencils"); pizza rollers, roller dockers, pastry wheels and bench scrapers (`kitchen-other`, a catch-all with no register); pasta machines with rollers (machines, not this leaf). One register covers the leaf; the `dough_job` split routes to the specialist pins.

Quote labels: **verbatim** = the sentence in quotation marks was in raw text I read myself (curl plus tag stripping, the Shopify product JSON, the eCFR XML, the Wikipedia plain-text API, or pdftotext on a PDF); **fetched excerpt** = the WebFetch tool returned or summarised the text; **search excerpt** = from a search-engine summary. This session's web-search budget was already used up by the time this run started (the WebSearch tool refused every query), so every source here was reached by fetching known URLs, Wayback snapshots, sitemaps and on-site search pages. No search excerpt is used anywhere.

## Seed

Seed output, recorded verbatim:

```
=== Example entry (from an unrelated category) ===
category: kitchen-knives
what this shows: a numeric spec that only makes sense as one side of a trade-off, and a visual grammar borrowed from the hobbyist community that puts the trade-off on one axis.

<example>
{
  "id": "steel_hardness",
  "label": "Steel hardness",
  "definition": "How hard the blade steel is, measured on the Rockwell C scale, which sets how long an edge stays sharp and how easily it chips.",
  "why_experts_care": "Harder steel holds a thinner, sharper edge for longer but is brittle, so it chips on bones and hard squash and is harder to sharpen at home. Softer steel dulls faster but bends rather than breaks and comes back with a few strokes on a honing rod. Japanese knives cluster at the hard end and German knives at the soft end, which is most of what the Japanese-versus-German debate is actually about.",
  "kind": "numeric",
  "unit": "HRC",
  "range": { "min": 52, "max": 66, "typical": "German-style knives sit around 56 to 58. Japanese-style knives sit around 60 to 63." },
  "roles": ["annotation", "sort_key"],
  "novice_answerable": false,
  "source": "listing",
  "listing_field_hint": "Often in the specs table as HRC or Rockwell. Missing on cheaper knives, which usually means soft.",
  "backfill_cost": "low",
  "backfill_note": "Manufacturer pages carry it for anything mid-range and up. Below that, infer soft from the steel name if given.",
  "visual": { "chartable": true, "grammar": "edge retention versus toughness scale", "encoding": "position on a single horizontal scale, hard on the right, with chip risk shaded in" },
  "priority": { "narrows_set": 3, "cost_to_answer": 4 },
  "evidence": ["knifesteelnerds_hardness", "seriouseats_knives"]
}
</example>

Trade-off entry it belongs to:
{
  "optimise": "steel_hardness",
  "give_up": "toughness and ease of sharpening",
  "plain": "Push hardness up and you get a knife that stays sharp for months but needs a whetstone and care around bones. Push it down and you get a knife you can abuse and hone in the kitchen but will sharpen more often.",
  "see_more_branch": "See more low-maintenance knives"
}

=== Persona ===
A rock climber, who thinks in failure modes and in what certification actually tested

Ask yourself two things and write the answers under 'Seed' in notes.md:
  1. What is the rolling-pins equivalent of the example above?
  2. What would this persona notice first about rolling-pins that a typical buyer would not?
Keep what survives the data check. Drop the rest.
```

1. **The rolling-pin equivalent of steel hardness is the pin's weight.** It is a number that only means something as one side of a trade-off: heavier pins do more of the work on cold, stiff dough, but past a point they dent dough when stopped, crush the air out of pizza dough and crack flaky crust. Two test kitchens put the sweet spot in the same place (ATK: favourites "from about 1 to 1½ pounds"; Wirecutter: picks at 23 and 24 ounces, lighter 11 to 16 ounce pins hard for a small tester; both verbatim). **Kept** as `weight_oz` with a one-axis band as the visual grammar. A second candidate surfaced on the way: the flat middle of a tapered pin, which trades pivoting against an even sheet. **Kept** as `flat_section_in`.
2. **The rock climber looks first at the joint, not the barrel, and at what "non-stick" and "food safe" were actually tested against.**
   - The joint: on a handled pin the axle is the failure point. ATK: "It's possible for the handles to turn instead of the barrel of the pin, which means that the pin just skids along a dough rather than flattening it" and "we worried about flour, dough, and water getting stuck in the space where the handles attach" (verbatim). Wirecutter: upgrade if you have "a handled pin that spins in place instead of rolling smoothly" (verbatim). Fletchers' Mill sells "genuine replacement handle sets" (verbatim). **Kept** as `bearing_type`, a low-priority annotation, with the honest note that no test compared bearing types.
   - "Non-stick": Zulay calls its marble pin "durable, non-stick marble" and its steel pin "Smooth, Non-Stick Surface" (verbatim, product JSON). ATK tested a Zulay steel pin and found "dough sticking to the barrel" and that "Marble, stainless-steel, and plastic models were also too smooth to hold a dusting of flour and stuck to pie dough" (verbatim). **Kept** as a judgement with `sources_disagree: true`.
   - "Food safe": nothing certifies rolling pins as a product, but the FDA Food Code names them: wood is allowed for "UTENSILS such as rolling pins" only if "Hard maple or an equivalently hard, close-grained wood" (verbatim, pdftotext). **Kept** as a judgement and as the reason `wood_species` exists.
   - **Dropped:** a "cracks or warps in the dishwasher" failure facet. Every maker says hand wash only, but no source documents a failure rate, so it stays as care advice in a judgement.
   - **Dropped:** ring wear or loosening on adjustable pins. ATK found flour trapped between ring and barrel, which is in `thickness_guides`; nothing on wear.

## Sources and what each supports

### Specialist reviewers who test

**Wirecutter, The Best Rolling Pin (Hannah Kirshner), read from the Internet Archive snapshot of 2025-11-11 because nytimes.com blocks fetches**
http://web.archive.org/web/20251111225858/https://www.nytimes.com/wirecutter/reviews/best-rolling-pin/
- verbatim: "When we tested 12 carefully selected rolling pins (out of nearly 30 we considered) on three types of dough"; "It's been years since writer Hannah Kirshner tested rolling pins for this guide, but we stand by her research and recommendations." So the testing is older than the snapshot; the page does not give the test year.
- verbatim: "most professional bakers and all of our experts use a handleless pin, because it gives you the best sense of contact with the dough and a better feel for if it's level."
- verbatim: "Handleless pins come in two styles: ones that taper towards the end, often referred to as French rolling pins, and straight dowels that are sometimes called shaker pins."; "Tapered pins pivot easily as you roll, making them ideal for flattening round pie crusts."; "sometimes it causes subtle hills and valleys, particularly if the taper is too extreme."; "Dowel pins have a slight advantage for creating large surfaces that are perfectly even. And they can be fitted with spacers (essentially plastic rings)".
- verbatim: "The ideal length of a handleless pin is 18-22 inches. If it's tapered, look for a straight section in the middle of seven inches or more, or look for a very gradual arc across the whole pin."
- verbatim: "light, narrow pins are more agile, while heavier pins do more of the work for you. A good pin shouldn't be so heavy that it cracks a flaky piecrust, or so light that you have to lean your whole weight into it".
- verbatim: "Maple and beech, the most common hardwoods used for rolling pins, provide good value, durability, and a pleasing weight. Some high-end rolling pins are made from other hardwoods, like walnut or cherry, while the cheapest pins use beechwood."
- verbatim: "Marble pins are beautiful but heavy. Aluminum pins can discolor eggy dough; neither metal nor glass hold flour well. Pins meant to be filled with ice water create condensation that can dampen what you're rolling out."; "conductive materials like marble and metal might be good at getting cold, but they are also good at getting warm quickly ... wood is a terrible conductor".
- verbatim: Whetstone Woodenware French pin "At 23 ounces"; "Lighter pins, in the 11-to-16-ounce range, proved difficult for someone of smaller stature"; "heavier non-tapered pins we tested left dents in the dough when not in constant motion"; Whetstone "had the longest straight section in the middle, measuring 12 inches compared to the typical six to eight".
- verbatim: JK Adams dowel "this 24-ounce dowel"; "A generic set of spacers (essentially rubber bands of various thicknesses), which won't work on a tapered pin".
- verbatim: JK Adams handled pin "can also be fitted with spacers like a dowel pin"; "The main advantage of a handled pin is that it does a lot of the work for you".
- verbatim: Vermont Rolling Pin cherry French pin "so slick that flour didn't adhere like it did to the other wooden pins"; "only 7 flat inches ... was too extreme for good handling or smooth rolling."
- verbatim: "If all you ever roll out is forgiving yeasted dough for pizza or sticky buns, a wine bottle might be all you need"; "the best rolling pin is one that you are comfortable with"; caption "Length is important for rolling big sheets of cookie dough."
- verbatim: "The higher-end pins were already treated with mineral oil or beeswax, so we rubbed mineral oil into the untreated wooden pins the night before testing".
- Supports: handleless over handled, tapered versus dowel, length, flat section, weight, materials, spacers. Confidence high (former pastry baker; three doughs; experts interviewed).

**America's Test Kitchen, The Best Rolling Pins (curl, live page)**
https://www.americastestkitchen.com/equipment_reviews/1381-rolling-pins
- verbatim: "We tested rolling pins in three styles: handled pins, dowels, and tapered pins (also called French rolling pins). The rolling pins in our lineup were made from marble, plastic, stainless steel, and wood."
- verbatim: "We didn't like rolling pins with handles. It's easier to assess a dough's thickness and how hard to press when using a dowel or tapered pin."; dowels' "heft and length make them especially suited for large jobs such as croissants or slab pies. Tapered pins are lighter, thinner, and offer finer control."
- verbatim: "The rolling pins we tested weighed from about 8 ounces to more than 4 pounds, a dramatic range. Our favorites weighed from about 1 to 1½ pounds".
- verbatim: "Most pie recipes call for a 12-inch circle of dough"; "Rolling pins with shorter barrels left unwanted indents where their edges pressed into the dough. These pins also required more passes over the dough ... This can also overwork and heat up the dough, causing it to turn out tough".
- verbatim: "The marble model we tested weighed 4.5 pounds, which felt cumbersome and crushed the air bubbles out of proofed pizza dough."; "Stainless steel in particular is sensitive to warmth as well as to cold; when we didn't pre-chill it, the pin warmed under our hands"; "Marble, stainless-steel, and plastic models were also too smooth to hold a dusting of flour and stuck to pie dough as a result."
- verbatim: "the heaviest pins crushed the air bubbles out of pizza dough while rolling."
- verbatim: "Although the tapered pins were roughly 20 inches long, each had different lengths of usable flat space, from 5.2 to 9.5 inches. It was easy to make unwanted indents while using the pins with the shortest amount of usable space".
- verbatim (handles): "It's possible for the handles to turn instead of the barrel of the pin, which means that the pin just skids along a dough rather than flattening it."; on a handled maple pin "we worried about flour, dough, and water getting stuck in the space where the handles attach."; OXO "flour and dough can get stuck in the nooks and crannies where the handles meet the pin."
- verbatim (rings): "One rolling pin had a 13.6-inch long barrel with sets of removable plastic rings ... The rings came in four sizes: 1/16-, 1/6-, 1/4-, and 3/8-inch."; "their hands ... landed directly on top of the rings, which pressed uncomfortably into our palms"; "the rings made deep impressions in doughs wider than 13.5 inches"; "flour was caught in between the barrel and the ring". Same pin: "We were able to roll out a 12-inch pie dough circle, but it left indents in the larger pizza and cookie doughs."
- verbatim (care): "Wash your rolling pin with soap and water and then dry it immediately and thoroughly. As with other wooden tools in the kitchen, don't submerge the pin in water. If you notice your pin beginning to feel dry, rub in a small amount of food-grade mineral oil."
- Per-model spec rows read from the page (verbatim): JK Adams dowel 1.4 lb, 19 in; JK Adams French 15.9 oz, 20.3 in, flat 9.5 in; Matfer beechwood dowel 1.3 lb, 19.4 in ("slightly smaller circumference required just a bit more effort ... a few more spins"); KitchenAid tapered 15.7 oz, 20.2 in, flat 8.5 in; a handled maple pin 2.8 lb, 14.7 in barrel ("the second-heaviest"); Ateco French 8.6 oz, 20 in, flat 5.2 in ("too dainty"); Joseph Joseph adjustable 14.1 oz without rings, 13.6 in; OXO nonstick handled 1.6 lb, 11.5 in ("flour didn't cling to the supersmooth surface and dough stuck"); Zulay stainless 9.7 oz, 15.6 in ("dough sticking to the barrel"); a Zulay 17-inch marble pin is in the lineup.
- The fields ATK reports per pin: Model Number, Style, Material, Weight, Barrel Length, and Flat Surface for tapered pins. That order is a tell for what testers compare.
- Supports: everything Wirecutter supports, independently, plus the weight band, the 12-inch pie circle, the flat-section range, the ring problems and the care routine. Confidence high.

### Regulators

**FDA Food Code 2022 (January 18, 2023 version), section 4-101.17 Wood, Use Limitation, and Annex 3 (PDF saved by WebFetch, then pdftotext)**
https://www.fda.gov/media/164194/download
- verbatim: "(A) Except as specified in ¶¶ (B), (C), and (D) of this section, wood and wood wicker may not be used as a FOOD-CONTACT SURFACE. (B) Hard maple or an equivalently hard, close-grained wood may be used for: (1) Cutting boards; cutting blocks; bakers' tables; and UTENSILS such as rolling pins, doughnut dowels, salad bowls, and chopsticks".
- verbatim (D)(2): treated wood containers must use "a preservative that meets the requirements specified in 21 CFR 178.3800 Preservatives for wood."
- verbatim Annex 3: "Moist foods may cause the wood surface to deteriorate and the surface may become difficult to clean. In addition, wood that is treated with preservatives may result in illness due to the migration of the preservative chemicals to the food; therefore, only specific preservatives are allowed."
- verbatim 4-101.11: utensil materials "may not allow the migration of deleterious substances or impart colors, odors, or tastes to FOOD".
- The Food Code is a model code for retail food businesses that states adopt; it does not bind home kitchens. It is still the regulator's own statement of which woods are fit for a rolling pin. Confidence high.

**21 CFR 172.878 White mineral oil, and 21 CFR 178.3620 Mineral oil (eCFR XML)**
https://www.ecfr.gov/current/title-21/section-172.878 and https://www.ecfr.gov/current/title-21/section-178.3620
- verbatim 172.878: "White mineral oil may be safely used in food in accordance with the following conditions"; use 5: "In bakery products, as a release agent and lubricant".
- verbatim 178.3620(a): "White mineral oil meeting the specifications prescribed in § 172.878 of this chapter may be used as a component of nonfood articles provided such use complies with any applicable limitations".
- Supports: what "food-grade mineral oil" means as a finish. Confidence high.

**21 CFR 177.2600 Rubber articles intended for repeated use (eCFR XML)**
https://www.ecfr.gov/current/title-21/section-177.2600
- verbatim: lists "Silicone (Si) elastomers containing methyl groups" among permitted elastomers; aqueous foods: "total extractives not to exceed 20 milligrams per square inch during the first 7 hours of extraction"; "(g) ... finished rubber articles intended for repeated use in contact with food shall be thoroughly cleansed prior to their first use".
- Supports: the rule silicone pins and silicone-covered barrels fall under. The extractive limits are not used in the JSON. Confidence high.

EU Regulation 1935/2004 (the EU food-contact framework) could not be read: EUR-Lex returned an empty body, the Commission page returned navigation only, the Wayback API returned 429, and legislation.gov.uk returned an empty page. Not cited.

### Brand spec pages (what fields brands publish)

**JK Adams (Shopify product JSON and product pages, curl)**
https://www.jkadams.com/collections/rolling-pins
- verbatim: Plain Dowel "a generous 19" surface area"; "slightly textured North American hardwood holds a dusting of flour better than any material"; size line "19" x 2"".
- verbatim: French Tapered variants "20 1/2" x 1 1/2"" and "20 1/2"x1 3/4"".
- verbatim: Gourmet (handled) "waxed hardwood barrel ... full-length steel axle and self-lubricating nylon bearings"; variants "12" x 2 3/8"" and "12" x 2 3/4"". Patisserie (handled) "precision steel ball bearings, a smooth waxed hardwood barrel".
- verbatim: Dumpling Dowel size line "9" x 1"" (the page title says "Maple Dumpling Dowel Wood Rolling Pin, 19"", which contradicts the body; the register uses only "about 1 inch thick" from it).
- verbatim: Lovely pin "Slight indentations on the edges of the pin ensure that your dough is rolled to a perfect, even thickness"; "the 1/4" thickness; this is the most common thickness for cookies, pie crusts and many other pastries."
- verbatim care: "Handwash with hot, soapy water after each use. Dry thoroughly with a towel. Never submerge in water. Lightly apply wood conditioner with a cotton cloth. Do not put in dishwasher, microwave or oven."
- Variant shipping weights in the product JSON: French Tapered "1814" g, "4.0" lb for a pin ATK weighed at 15.9 oz; Plain Dowel "907" g, "2.0" lb for a pin ATK weighed at 1.4 lb and Wirecutter at 24 oz. This is why `weight_oz` is researched.
- Fields: size as length x diameter in the variant name, wood ("North American hardwood"), bearing type for handled pins, care. No weight, no flat-section length.

**Fletchers' Mill (Shopify product JSON, curl)**
https://www.fletchersmill.com/
- verbatim: 15" pin "Crafted from solid rock maple ... Precision steel ball bearings allow the barrel to glide easily"; variants "Untreated/natural" and "Finished".
- verbatim: "12", 15", 18" Rolling Pin Handle Replacement Set"; "Restore the look and functionality of your Fletchers' Mill rolling pin with genuine replacement handle sets".
- verbatim: "Our 8" Tortilla Rolling Pin is perfect for rolling out tortillas, flatbreads"; "20″ French Rolling Pin ... Tapered from the center to the ends"; "18.5" Bakery Rolling Pin ... 3.5" in diameter".
- Fields: barrel length in the title, wood species, bearing type, finished or untreated. Handled pins are named by barrel length (10, 12, 15, 18 in).

**Vermont Rolling Pins (product pages, curl)**
https://www.vermontrollingpins.com/
- verbatim: Long Pasta/Mattarello "Approx. 1 3/4" in diameter and approx. 36" to 40" in length."
- verbatim: Belan "the traditional rolling utensil to make many Indian breads like Roti, Chapati, and Naan"; "Approx. 2 " in diameter and approx. 13" in length".
- verbatim: Oklava "a Turkish rolling pin used to make Gozleme ... The cylinder is super thin"; "Approx. 1 " in diameter and approx. 32" in length."
- verbatim: Lefse "hand turned with the prerequisite corrugation"; Tutove "the vertical ridges on this pin allow for the sheets of butter to be incorporated into the pastry without breaking it."
- verbatim: Cookie pin "either a built in 1/4 inch or 3/16" guide at the handles"; French "Approx. 1 3/4" in diameter and approx. 20" in length"; Shaker "Approx. 2 ¾" in diameter and approx. 20" in length".
- verbatim care: "After use wipe with a damp cloth. Do not use soap, it will dry out your wood."; "Oil with a food safe oil".
- verbatim woods: "Cherry has a tight grain and is resistant to warping."; "Hard maple resists abrasion; it is strong and heavy."; "Walnut has a hard and tightly dense grain".
- verbatim FAQ: "Pictures can be deceiving and without dimensions, you could be purchasing a smaller rolling pin than you thought."
- Supports: the specialist pin families and their sizes, the soap disagreement, wood species notes. Confidence high for what the maker states.

**King Arthur Baking, French Rolling Pin product page (curl)**
https://www.kingarthurbaking.com/shop/items/french-rolling-pin
- verbatim: "Unfinished wood takes on flour over time, becoming less sticky than finished pins."; "this patisserie-style French rolling pin allows a greater "feel" for the dough".
- Supports: unfinished over finished. Confidence medium (retailer, but a baking company).

**King Arthur Baking, Classic Double Pie Crust recipe (WebFetch)**
https://www.kingarthurbaking.com/recipes/classic-double-pie-crust-recipe
- fetched excerpt (quoted by the tool): "If your pan is 7" across the bottom, and 1 1/2" up each side, that's a total of 10". This means you should roll your bottom crust to a diameter of about 12"".
- Supports: second source for the 12-inch pie round. Confidence high.

**Joseph Joseph, Adjustable Rolling Pin (curl)**
https://www.josephjoseph.com/products/adjustable-rolling-pin
- verbatim: "five sets of removable, colour-coded rings that allow you to raise the rolling surface to 2mm (1/16"), 4mm (1/6"), 6mm (1/4"), 8mm (5/16") or 10mm (3/8")"; "High quality beech pin"; "Rings: Hand wash only. Wooden pin: Hand wash only. Do not submerge in water and dry immediately after washing."; "H6.4 x W42 x D6.4cm".
- ATK tested a four-ring version (1/16 to 3/8 in); the current page lists five. Same span.
- Supports: ring thicknesses (with ATK), hand-wash rule. Confidence high.

**Wilton, Fondant Roller with Guide Rings, 20-Inch (curl, found through the site's XML sitemap)**
https://wilton.com/fondant-roller-with-guide-rings-20-inch-fondant-tools/191000329/
- verbatim: "great for rolling out large amounts of fondant for covering tiered cakes"; "three pairs of guide rings ... gold rings 3/16 in. (4.8 cm), orange rings 1/8 in. (3.2 cm) and blue rings 1/16 in. (1.6 cm)" (the metric figures are wrong on the page: 3/16 in is 4.8 mm, not cm); "Material: Plastic roller; rubber rings"; "20 in. (50.8 cm) long"; "The fondant roller's non-stick surface ... just dust the surface with icing sugar".
- Supports: the fondant family (plastic, about 20 in, guide rings, icing sugar). Confidence high for the product; one brand.

**Zulay Kitchen (Shopify product JSON, curl)**
https://zulaykitchen.com/
- verbatim: marble pin "Heavy marble design provides natural weight"; "Crafted from durable, non-stick marble"; steel pin "Smooth, Non-Stick Surface: Keeps dough from sticking".
- Supports: the brand side of the non-stick disagreement. ATK tested Zulay's steel pin.

### Primers and recipe sites

**Wikipedia, Rolling pin (plain-text API)**
https://en.wikipedia.org/wiki/Rolling_pin
- verbatim: "Rod: thin rods typically made of wood around 1–2 inches (2–3 cm) in diameter. They are used by rolling the rod across the dough using one's palm. The pins may be tapered at one or both ends for more pivot control in certain tasks such as making small jiaozi skins or pie shells. Most East Asian or French style rolling pins, and the Turkish Oklava are rod-style."
- verbatim: "Roller: consists of a thick heavy roller made of a variety of materials around 3–4 inches (7–10 cm) in diameter with thinner handles which extend through the roller."
- verbatim: "Textured: some specialized rolling pins have textured surfaces that mark and indents the dough surfaces"; "Some are hollow and are able to be filled with cold or warm water".
- Supports: rod versus roller, textured pins, dumpling pins are thin rods. Confidence medium.

**Italian Wikipedia, Matterello (plain-text API)**
https://it.wikipedia.org/wiki/Matterello
- verbatim: "un semplice cilindro in legno di 5-8 cm di diametro e lungo 50-70 cm (90-130 cm in Emilia e in Romagna)" (a simple wooden cylinder 5 to 8 cm across and 50 to 70 cm long, 90 to 130 cm in Emilia-Romagna).
- Supports: the pasta pin is a long plain cylinder; second source with Vermont for "about 90 cm or longer". The 5 to 8 cm diameter disagrees with Vermont's 1 3/4 in (4.4 cm); not used.

**The Woks of Life, Dumpling Wrapper Recipe (WebFetch; curl returned 403)**
https://thewoksoflife.com/dumpling-wrapper-recipe/
- fetched excerpt (quoted by the tool): "A small Chinese rolling pin is the best tool for rolling out dumpling wrappers. It's small (only about 11 inches/28cm long and at most, 1 inch/2.5cm thick), making it easy to maneuver with the small pieces of dough."; "roll the dough out into a 3.5 inch (9cm) circle".
- Supports: the dumpling pin's size. Confidence medium-high.

**China Sichuan Food, Dumpling Wrappers (WebFetch)**
https://www.chinasichuanfood.com/dumpling-wrappers/
- fetched excerpt (quoted by the tool): "Rotate the wrapper when moving the rolling pin. The wrappers will have thicker centers and thinner edges naturally. The wrapper should be around 8cm to 10cm in diameter."
- Supports: dumpling wrappers are not rolled to one even thickness, which is why thickness guides drop out on that branch. Confidence medium.

### Could not reach

Serious Eats (HTTP 402 on every URL; no Wayback snapshot found for the guessed URLs), The Kitchn (403), Red House Spice (403), Food & Wine and King Arthur blog guesses (404, no snapshot), Whetstone Woodenware (no product JSON), EUR-Lex and the Commission food-contact page (empty or navigation only), the Wayback CDX index for several domains (504 or 429). Amazon and other retail listings were not attempted; prior registers record that they block fetches.

## Data points used for derived metrics

- `barrel_margin_in` = barrel length minus the width of the dough. Pie target 12 in: ATK ("Most pie recipes call for a 12-inch circle", verbatim) and King Arthur ("about 12"", fetched excerpt). Why the margin matters: ATK, short barrels "left unwanted indents where their edges pressed into the dough" and needed more passes (verbatim); the 13.6 in Joseph Joseph pin rolled a 12-in pie round but marked larger doughs (verbatim). The register gives no minimum margin, because no source states one.

## Numbers in category.json and how many sources stand behind each

| Number | Sources |
|---|---|
| Handleless pin length about 19 to 22 in (Wirecutter 18 to 22; ATK favourites 19 to 20.3) | Wirecutter, ATK |
| Weight sweet spot about 1 to 1.5 lb (16 to 24 oz) | ATK ("about 1 to 1½ pounds"); Wirecutter (picks 23 and 24 oz) |
| 11 to 16 oz pins hard for a small tester | Wirecutter; consistent with ATK's 8.6 oz Ateco "too dainty" |
| Marble pins heavy; ATK's weighed 4.5 lb | ATK (number); Wirecutter ("beautiful but heavy", no number). The number is attributed to ATK in the text. |
| Flat section: ATK 5.2 to 9.5 in on about 20 in pins; Wirecutter typical 6 to 8, its pick 12 | Each figure is that tester's own measurement, attributed in the text; the claim they support (short flat sections roll unevenly) has both |
| Pie round about 12 in | ATK, King Arthur |
| Ring and guide thicknesses span about 1/16 to 3/8 in | Joseph Joseph, ATK; Wilton 1/16 to 3/16 within it |
| 1/4 in is the common cookie and pie thickness | JK Adams (verbatim); Vermont cookie pin guides at 1/4 or 3/16 in |
| Handled pins mostly 10 to 15 in barrels | Fletchers' Mill (10, 12, 15, also 18), JK Adams (12), ATK (11.5, 14.7) |
| Dumpling and small pins about 1 in thick, under 12 in long | Woks of Life (11 in, at most 1 in); JK Adams dumpling dowel (9 x 1 in); Fletchers' tortilla pin (8 in); Wikipedia (rods 1 to 2 in) |
| Pasta pin about 90 cm (36 in) or longer | Vermont (36 to 40 in), Italian Wikipedia (90 to 130 cm in Emilia-Romagna) |
| Fondant roller about 20 in, plastic | Wilton (20 in); one brand, stated as Wilton's figure |

## Against the existing add-on schema

Add-on read: `supplies/kitchen/add-ons/rolling_pin.jsonc`, which is empty (`{}`), composed with `kitchen-standard.jsonc` (`materials[]` with roles, `max_heat_temp`, `dishwasher_safe`) and `supplies-standard.jsonc` (`high_signal_markdown`, `sizing` with `dimensions`, `weight`, `pack_count`, `pieces_included_text`). Kinds follow AGENTS.md: vertical anchor, category quality currency, hard filter, sizing, or markdown. Proposals only; nothing in zamana-expertise was edited.

| Facet in this doc | In schema today | Proposed kind under AGENTS.md | Note |
|---|---|---|---|
| dough_job | no | not stored; derived at query time | Comes from pin_style, barrel_material and barrel length. Listing "ideal for pie, pizza, fondant" bullets are marketing and list every dough. |
| pin_style | no | none (image pipeline), plus title words | Handles versus handleless and tapered versus straight are answered by the photo, so rule 1 keeps them out. The title usually says "French", "tapered", "dowel" or "handled", which the text pass can keep. Small and pasta pins fall out of barrel length; textured pins are a photo fact. |
| barrel_material | yes, `materials[]` role `body` (wood, marble, stainless_steel, aluminum, plastic, acrylic, silicone; `ptfe_nonstick` as `coating`) | vertical anchor (existing) | Already covered. Make sure handled pins also record the handle material with role `handle`, and ring sets as `other`. |
| wood_species | partly: the enum has `wood` only | high_signal_markdown | The Food Code test is "hard maple or an equivalently hard, close-grained wood". One markdown line ("solid rock maple", "beechwood") carries it. Not worth an enum. |
| wood_finish | no | high_signal_markdown | "Untreated", "oiled with mineral oil", "waxed". Fletchers' Mill sells the same pin both ways, so the line belongs with the variant text if the PDP offers it. |
| barrel_length_in | partly: `sizing.dimensions.length`, but it does not say whether that is the barrel or the whole pin | sizing (propose `sizing.rolling_length` with unit) | The category's bought-by-number dimension, and it varies by variant (JK Adams French in two sizes; Fletchers' 10 to 18 in). Trap comment needed: handled pins are named by barrel length, while some listings give overall length including handles. |
| barrel diameter | no; `dimensions` has no diameter | high_signal_markdown | Brands print "L x D" ("19" x 2""). Reviewers call diameter subjective (Wirecutter). Not a filter. |
| flat_section_in | no | none (researched, outside schema) | Only testers measure it. Top-N backfill from reviews at most. |
| weight_oz | yes, `sizing.weight` | sizing (existing), with a trap comment | Trap: the store "weight" on JK Adams is the shipping weight (4.0 lb listed, 15.9 oz measured by ATK). Only fill from a stated product weight or a review. It is the one number we would augment, so it qualifies as quality currency. |
| thickness_guides | no | hard filter (enum) | Proposal: `thickness_guides` enum [none, removable_rings, built_in_guides, separate_spacers, unspecified]. "Adjustable rolling pin" is a shopping term in its own right, and the photo does not always show rings. Ring thicknesses go in `pieces_included_text`. Borderline under rule 5 (feature vocabulary); kept because buyers filter by it. |
| bearing_type | no | high_signal_markdown | "Steel ball bearings", "nylon bearings on a steel axle". No test ranks them. |
| dishwasher_safe (kitchen-standard) | yes | keep | Almost always `no` for wood; `yes` on some plastic and silicone pins. Carries the label, while the judgement carries the advice. |

Add-on and standard fields this register considers noise for this leaf: `max_heat_temp` (no pin meets heat; leave null; one maker says "Do not put in ... microwave or oven", which is care text); `sizing.capacity` (only hollow fill-with-water pins have one, and reviewers advise against them); `sizing.dimensions.width`, `height` and `depth` beyond length and diameter (a pin is a cylinder). `pack_count` is fine for pin-and-ring sets but carries no expertise.

## Branches

Every leg below traces to a sentence above. Chains are written out where a leg needs two facts. Updated 2026-10-25 after the blind check; changed legs are marked.

`dough_job`
- **pie and pastry → ask pin_style, weight_oz.** Wirecutter: tapered pins "ideal for flattening round pie crusts"; weight: "shouldn't be so heavy that it cracks a flaky piecrust" (Wirecutter). SUPPORTED. **Changed 2026-10-25**: dropped `flat_section_in` from this leg. It needs `pin_style = tapered French` first, and that branch already asks it, so asking it here was redundant and unanswerable before the style question runs (check.md Part C, pie → flat_section_in).
- **cookies → ask thickness_guides, barrel_length_in.** JK Adams: 1/4 in is "the most common thickness for cookies"; Vermont makes a cookie pin with a built-in guide; Wirecutter caption: "Length is important for rolling big sheets of cookie dough"; ATK: a 13.6 in barrel "left indents in the larger pizza and cookie doughs". SUPPORTED.
- **fondant → ask barrel_material, thickness_guides.** Wilton's fondant roller is plastic with a non-stick surface dusted with icing sugar, and ships with 1/16 to 3/16 in guide rings. SUPPORTED, but on one brand's product; labelled as such in the register.
- **dumplings and flatbreads → ask barrel_length_in; bearing_type irrelevant.** Length: the dumpling pin is defined by being short (Woks of Life, about 11 in; JK Adams 9 x 1 in; Fletchers' 8 in tortilla pin), though the thicker, longer belan (Vermont: 2 in by 13 in) also belongs here by name, not by this length rule (see Corrections #1). Bearings: chain of two facts, (1) dumpling pins are thin rods (Wikipedia: rods are tapered "for more pivot control in certain tasks such as making small jiaozi skins"), (2) rods are rolled "using one's palm" with no handles (Wikipedia), so there is no bearing. SUPPORTED. **Changed 2026-10-25**: dropped `thickness_guides` from makes_irrelevant. The chain (a guide holds one thickness; wrappers "will have thicker centers and thinner edges naturally", China Sichuan Food) is sourced for dumpling wrappers only; nothing cited says roti, chapati or tortillas roll unevenly or that guides fail on them (check.md Part C, WEAK). Left as a leg for dumpling wrappers specifically would need splitting the facet value, which is out of scope for this pass; dropping the irrelevance claim is the safer fix.
- **fresh pasta → ask barrel_length_in; bearing_type and flat_section_in irrelevant.** Length: 36 to 40 in (Vermont), 90 to 130 cm in Emilia-Romagna (Italian Wikipedia); an everyday mattarello elsewhere in Italy is only 50 to 70 cm (20 to 28 in), see Corrections #1. Irrelevant: Italian Wikipedia describes it as "un semplice cilindro in legno", a plain cylinder, so there is no taper and no handle. SUPPORTED.
- **pizza and bread → ask weight_oz.** ATK: "the heaviest pins crushed the air bubbles out of pizza dough"; Wirecutter: forgiving yeasted dough needs little more than "a wine bottle". SUPPORTED. The value stays in `dough_job.values` and this branch for filtering, even though the cold-start question no longer lists "pizza" as a menu word (Corrections #13).

`pin_style`
- **handled roller → ask bearing_type, barrel_length_in; flat_section_in irrelevant.** The joint is the handled pin's failure point (ATK skidding and trapped flour; Wirecutter "spins in place"); handled pins run 10 to 15 in barrels (Fletchers', JK Adams, ATK) so length varies most here; flat section is defined only for tapered pins. Thickness guides stay relevant: Wirecutter says the handled pin "can also be fitted with spacers like a dowel pin". SUPPORTED.
- **tapered French → ask flat_section_in, weight_oz; thickness_guides and bearing_type irrelevant.** Flat section: Wirecutter and ATK. Weight: tapered pins are the light ones (ATK "Tapered pins are lighter, thinner"; Ateco at 8.6 oz "too dainty"; Wirecutter 11 to 16 oz French pins hard for a 10-year-old tester). Guides: Wirecutter, spacers "won't work on a tapered pin". Bearings: no handles. SUPPORTED. This is the only leg that reaches `flat_section_in`, so pie buyers get it after answering `pin_style`, not before (see the pie leg above).
- **straight dowel → ask thickness_guides, barrel_length_in; flat_section_in and bearing_type irrelevant.** Wirecutter: dowels "can be fitted with spacers"; ATK: dowel "heft and length make them especially suited for large jobs". The whole barrel is flat and there are no handles. SUPPORTED.

`barrel_material`
- **wood → ask wood_finish, wood_species.** King Arthur: unfinished wood "becoming less sticky than finished pins"; Food Code: "Hard maple or an equivalently hard, close-grained wood". SUPPORTED.
- **marble → ask weight_oz; wood_finish and wood_species irrelevant.** ATK's marble pin "weighed 4.6 pounds [see Corrections #14], which felt cumbersome"; Wirecutter "Marble pins are beautiful but heavy". The wood facets do not apply to stone. SUPPORTED.
- **metal → wood_finish and wood_species irrelevant.** Definitional. SUPPORTED. A "metal → ask weight" leg was drafted with marble and dropped: ATK's Zulay steel pin weighed only 9.7 oz, so nothing says steel pins are heavy.
- **plastic, silicone or nonstick coating → wood_finish and wood_species irrelevant.** Definitional. SUPPORTED. (Metal and plastic share this leg in the JSON, as one branch.) **Changed 2026-10-25**: a nonstick-coated or silicone-covered steel barrel (ATK's OXO) now explicitly files as "silicone or nonstick coating", not "metal", in `barrel_material`'s backfill_note, since the coating is what changes how it performs (check.md Part B, split 3c).

`skip_if_known` on `pin_style` skips the style question when the dough answer is dumplings, pasta or fondant: the first two have their own pin families (small thin pin, long pasta pin), and the only fondant roller read (Wilton) is a straight plastic cylinder, so the material question does the work there.

Dropped legs: "fondant → wood_finish irrelevant" (a wood pin can roll fondant; no source says it cannot); "pasta → thickness_guides irrelevant" (no source); "cookies → flat_section_in irrelevant" (Wirecutter's tapered winner "works well for rolling both pies and cookies").

**Changed 2026-10-25**: the `barrel_material` cold-start question changed from "Wood, a chilled marble or steel pin, or smooth plastic?" to "Wood, marble, steel, or plastic?" so each answer maps to exactly one facet value (marble and metal used to share one answer option even though their branches differ — check.md Part B, split 3c).

## Trimmed detail

- `pin_style`: Wikipedia lists a "Hybrid" style with tapered ends and ball bearings (its example is a brand name). Too rare to be a value; it maps to handled roller.
- `pin_style`: Vermont's tutove (ridged, for puff pastry butter sheets) and lefse (corrugated) pins sit under "textured or embossed" with the decorative embossed cookie pins, though their purpose differs (working the dough versus printing a pattern).
- `barrel_material`: Wirecutter says aluminium "can discolor eggy dough" and that ice-water pins "create condensation that can dampen what you're rolling out". Wikipedia says marble pins "are often cooled in a refrigerator for maintaining a cold dough while making puff pastry".
- `weight_oz`: ATK's tested range was "about 8 ounces to more than 4 pounds". Not used as a facet range because only one source gives it.
- `flat_section_in`: Wirecutter recommends "seven inches or more", yet calls the 7-inch Vermont cherry pin "too extreme". Its torpedo shape may be the real fault, so the register states the direction (longer is more even) and not a floor.
- `barrel_length_in`: ATK's Matfer dowel had a "slightly smaller circumference" and "required just a bit more effort ... and a few more spins". That is the only tested statement on diameter, so diameter is not a facet.
- `wood_species`: Wirecutter both praises beech for "good value" and says "the cheapest pins use beechwood"; ATK's runner-up dowel is beech. Beech is fine.
- Care: Vermont's "Do not use soap" against JK Adams' "Handwash with hot, soapy water" and ATK's "soap and water". Everyone agrees on no soaking and no dishwasher.

## Corrections after the blind check (2026-10-25)

Scope addendum: pin-plus-cookie-cutter gift sets (JK Adams sells four) stay in `rolling-pins`, since the pin is the primary item and the cutters are a bundled extra; chakla-belan sets go to `kitchen-boards`, since the board (chakla) is the item the set is named and priced around, with the belan noted as an inclusion. (check.md Required fix 14, boundary cases from Part B split 1e.)

Sources: all WebFetch/pdftotext fetches below were run on 2026-10-25 from this run's own scratch dir. Web search was unavailable again (budget shared across concurrent agents, same as the original run), so every new claim below comes from fetching a known URL directly, same method as the original run.

1. **Fixed the specialist-pin sizes and the backfill thresholds (check.md's one CONTRADICTED finding).** Re-read Vermont's own belan (2 in diameter, 13 in long) and oklava (32 in) pages and the Italian Wikipedia matterello page (50–70 cm ordinary, 90–130 cm only in Emilia-Romagna) — all already cited, just misapplied. Rewrote the "Small thin pin" and "Long pasta pin" mechanism families to state the belan's real size instead of forcing it into "about 1 inch thick"/"under a foot", and to say the everyday mattarello (20–28 in) is shorter than the long style Vermont sells (32–40+ in). Rewrote `barrel_length_in`'s `why_experts_care` and `range.typical` to match (dumpling/flatbread pins 8–13 in; long pasta pins 32 in and up; everyday mattarello 20–28 in noted separately). Rewrote the `dough_job` backfill_note to check title keywords (dumpling, tortilla, roti, chapati, belan, mattarello, pasta, oklava) **before** falling back to a length rule, and narrowed the length fallback to under 13 in / 30 in and up, flagging the 13–30 in band as ambiguous and keyword-dependent. Rewrote the "Mattarello" vocabulary entry to say it is the generic Italian word for a rolling pin, not only the name of the long pasta pin.
2. **Restated `weight_oz` sourcing and dropped the shipping-weight rule.** check.md's own Amazon spot-check (not re-run by me; I could not get Amazon's rendered weight field through WebFetch, which returned only script/nav HTML) found store weights wrong both high and low, not just inflated, so the "reject above 3 lb" cutoff caught none of the real errors. Removed that cutoff from `weight_oz.backfill_note` and reworded `why_experts_care` and `listing_field_hint` to say the field is "noisy in both directions" rather than "often a shipping weight". Downgraded judgement 4's confidence from medium to low, set `sources_disagree: true`, and reworded its claim so it no longer asserts a one-directional "shipping weight" story.
3. **Marked the weight band's low edge as disputed.** Set `sources_disagree: true` on judgement 3 ("Pins of about 1 to 1.5 pounds suit most bakers…") and rewrote its claim and note to say ATK's own Best Tapered (15.9 oz) and recommended KitchenAid (15.7 oz) picks sit inside Wirecutter's 11-to-16-ounce "hard for a 10-year-old tester" range. Changed "a small tester" to "a 10-year-old tester" in `weight_oz.why_experts_care` and in the `wirecutter_pins` reference's supports line, matching check.md's reading of the source (Wirecutter names the tester as 10-year-old Vivi; I could not re-fetch the Wirecutter page myself this run — nytimes.com and the Wayback snapshot both refused my fetches with a bot-block page — so this rests on the checker's independent read confirming the register's own earlier verbatim quote, "someone of smaller stature", from the same page).
4. **Fixed the handleless length band.** Changed `barrel_length_in.why_experts_care` and `range.typical` from "19 to 22 in" to "18 to 23 in" to match Wirecutter's stated "18-22 inches" plus its 23-inch top pick, rather than a figure matching neither source.
5. **Reclassified `barrel_material` and `barrel_length_in` as `source: "listing"`** (from `researched`). Added trap language to both `listing_field_hint`s: a retail Material row can name the wrong wood species (check.md found "Oak" on a pin the brand page calls beech), and a title length can disagree with the dimension row (check.md found a listing titled "22-Inch" whose dimension row and the tester's own measurement both read 20.2 in). I could not re-verify Amazon's rendered spec rows myself (WebFetch returned only Amazon's script/nav shell, not the rendered product-details table), so this reclassification rests on check.md's own methodology section, which states it fetched these Amazon pages live with curl.
6. **Corrected "No maker lists it" for `flat_section_in`.** Vermont states a "barrel rolling area" figure for some pins (already cited as `vermont_pins`, now added to this facet's `evidence`). Reworded `why_experts_care` and `listing_field_hint` to say a maker's figure exists but can disagree with a reviewer's measurement, and kept the facet `researched` rather than `listing`, since the register treats the reviewer's number as the trustworthy one.
7. **Fixed the `pin_style` keyword map.** Reworded `listing_field_hint` and `backfill_note` to drop "Shaker" as a reliable dowel cue (Vermont's own Shaker pin has a 12 in rolling area on a 20 in pin, meaning integral handles) and to say a handled pin may be titled "Gourmet" or "Patisserie" with no "handled" word, and to give fixed-handle pins a fallback rule.
8. **Made the `barrel_material` question routable.** Changed the cold-start question from "Wood, a chilled marble or steel pin, or smooth plastic?" to "Wood, marble, steel, or plastic?" so each answer maps to one facet value instead of merging marble and steel into one answer whose branches differ. Added a rule to `backfill_note`: a nonstick-coated or silicone-covered steel barrel files as "silicone or nonstick coating", not "metal".
9. **Weakened the dumplings-and-flatbreads leg.** Dropped `thickness_guides` from that branch's `makes_irrelevant`, since the cited reasoning (guides fight the wrapper's natural thick-center taper) is sourced for dumpling wrappers only and not for roti, chapati or tortillas. Did not split the `dumplings and flatbreads` facet value itself — that is a larger structural change than this pass's scope, and check.md offered dropping the leg as an explicit alternative to splitting.
10. **Dropped `flat_section_in` from the pie branch.** It needs `pin_style = tapered French` first, which is a step the pie branch never states, and the tapered-French branch already asks it.
11. **Restricted `barrel_margin_in`'s scope.** Reworded its `why` to say it is the pie check for a straight or handled barrel, that a tapered pin should use `flat_section_in` instead, and that no dough besides pie has a stated target width, rather than implying the metric applies across all `dough_job` values it `depends_on`.
12. **Fixed the safety wording.**
    - Removed "Wood preservatives must meet 21 CFR 178.3800" from the `fda_food_code` reference's `supports` line. I re-fetched the FDA PDF myself (it loaded for me today, unlike the checker's run) and extracted it with pdftotext: §4-101.17(D)(2) ties 178.3800 only to treated wood containers for whole raw produce, not to rolling pins under (B). Replaced the line with the model-code point instead.
    - Changed "It binds food businesses, not homes" to "a model code that binds food businesses where adopted" in judgement 6's note and added the same point to the `fda_food_code` reference's own `supports` line.
    - Added a new judgement: cherry and walnut may not meet the Food Code's "equivalently hard, close-grained" test that hard maple meets. Sourced to two independent Janka-hardness tables I fetched this run, The Wood Database and Wikipedia's Janka hardness test page, which agree exactly: hard maple 1450 lbf, black walnut 1010 lbf, black cherry 950 lbf — both about two-thirds of maple's hardness. Added both as new references (`wood_database_hardness`, `wiki_janka_hardness`) and reworded `wood_species.why_experts_care` to flag cherry and walnut as unresolved rather than compliant.
    - Did **not** switch the `fda_food_code` citation to the Internet Archive copy, against check.md's suggestion. I fetched `https://www.fda.gov/media/164194/download` myself today and it returned the real PDF (4.8 MB, readable with pdftotext) rather than the abuse-detection page the checker hit, so the live FDA URL is citable again as of this run. Kept the live URL; noted the checker's access problem may be transient or client-specific rather than a change on FDA's side.
13. **Fixed the `dough_job` question.** It offered six answers, more than the two-to-four STYLE.md guideline, and "dumplings" alone gave no word for a roti or chapati buyer. Changed the question from "Pie, cookies, fondant, dumplings, pasta or pizza?" to "Pie, cookies, fondant, roti or dumplings, or pasta?": five grouped answers instead of six, with "roti" now named directly, and "pizza" dropped from the menu (per Wirecutter, forgiving pizza dough barely needs a pin) while the `pizza and bread` facet value and its branch stay for buyers who do name it or search for it. Added a sentence to `dough_job.definition` telling the assistant to ask which dough a multi-dough buyer rolls most, and that a product can carry more than one value, since the register already notes general pins serve several doughs but previously gave no rule for that.
14. **Minor fixes.**
    - Checked ATK's page myself: the narrative and the spec table both now read 4.6 lb for the marble pin, with "4.5 pounds" appearing once elsewhere on the same page. That number is not actually written into `category.json` anywhere (only into these notes), so no JSON edit was needed; recorded here so a future `weight_oz` backfill from ATK uses 4.6 lb.
    - Softened the handled-roller trade-off from "you cannot feel the dough's thickness" to "it's harder to feel the dough's thickness", matching ATK's own "more difficult to assess" framing rather than an absolute.
    - Resolved the pin-plus-cutter / chakla-belan boundary call above.

No finding from check.md was rejected outright; item 12's Internet Archive suggestion was the one place I diverged, for the reason given above (the live source is reachable again).

## Corrections On the first validate run (about 13:44) schema.json's `vertical` enum was [formulations, machines, fashion, furnishing, other] and validate.py reported one error for it. By 13:45 the schema had been changed outside this run to add "supplies" (I did not touch the skill), and validate, render and viewer then ran clean. Registers written by concurrent runs before that change may carry "other".
2. **Web search was unavailable for the whole run.** The WebSearch tool refused every call ("this session has used its web search budget (200 of 200 WebSearch calls)"), because concurrent agents share the budget. Sources were found by fetching known URLs, the Wayback availability API and CDX index, brand Shopify JSON, Wilton's XML sitemap, and the on-site search pages of two recipe sites. Serious Eats, the usual third reviewer, could not be reached. A rerun with search might add Serious Eats and a cake-decorating reviewer for the fondant branch.