# Measuring cups and spoons: research notes

Date: 2026-10-25. Scope: kitchen volume measures sold under `supplies > kitchen > measuring_cups`: nesting dry measuring cups, liquid measuring cups (spouted jugs in glass, plastic or silicone), measuring spoon sets, combined cup-and-spoon sets, and adjustable plunger measures for sticky ingredients.

Not in scope, left to siblings in `supplies > kitchen` (catalog.json):
- Jiggers and the OXO Mini Angled Measuring Cup when it is listed as barware go to `cocktail-mixing-tools`, whose notes already claim that overlap. A plain kitchen measuring cup stays here.
- Standalone measuring scoops and lone coffee scoops are claimed here, not sent to `manual-coffee-tea-prep`: that register's notes and JSON never mention scoops, and `scoops/notes.md` sends "measuring scoops and spoons" back to `measuring-cups`. `measure_type`'s definition states this ("a lone measuring or coffee scoop counts with spoons"). A scoop bundled inside a measuring set (Farberware 5-piece "with Coffee Scoop") stays here as one piece of the set either way.
- Mixing bowls with printed graduations go to `mixing-bowls`. Ice cream and cookie scoops go to `scoops`. A "25PCS" bowl set padded with measuring cups and spoons that `mixing-bowls` sends here is claimed too: `measure_type`'s backfill_note classes it by its measures, usually `cups and spoons set`.
- Kitchen scales are not a leaf of `supplies > kitchen` in catalog.json. They appear here only as the tool every source says beats volume measures.
- Pool, lawn and chemical measuring pitchers turn up in the Amazon "liquid measuring cup" search (B08WWY8KQK, B08PDS3PB2). They are not kitchen tools and are out of scope.

Quote labels: **verbatim** = the sentence was read in raw text from curl (HTML stripped of tags) or pdftotext, and is quoted exactly; **fetched excerpt** = the WebFetch tool summarised or reworded; **search excerpt** = from a search-engine summary only. This run had no web search budget left (the session's 200 searches were spent by concurrent runs), so nothing here is a search excerpt. Every quote below is verbatim from raw text unless marked otherwise.

## Seed

Seed output, verbatim:

```
=== Example entry (from an unrelated category) ===
category: car-tyres
what this shows: a number that is on every listing and still needs a judgement attached, because experts know it only compares within one manufacturer's range.

<example>
{
  "id": "treadwear_grade",
  "label": "Treadwear grade",
  "definition": "A number moulded on the tyre's side, from a US test, that estimates how long the tread lasts relative to a reference tyre set at 100.",
  "why_experts_care": "A 600 should last roughly twice as long as a 300, which is the single biggest driver of cost per mile. The catch experts repeat is that each manufacturer runs its own test, so the number is reliable for comparing tyres within one brand and loose across brands. Softer, grippier tyres sit low on this scale on purpose.",
  "kind": "numeric",
  "unit": "grade",
  "range": { "min": 60, "max": 1000, "typical": "Most everyday tyres sit between 300 and 700." },
  "roles": ["annotation", "sort_key"],
  "novice_answerable": false,
  "source": "listing",
  "listing_field_hint": "Listed as UTQG, usually as three values together, for example 500 A A. Treadwear is the first.",
  "backfill_cost": "low",
  "backfill_note": "If absent from the listing, the manufacturer's spec sheet always has it.",
  "visual": { "chartable": true, "encoding": "position on a horizontal scale, with the grip grade as a second mark so the trade-off shows" },
  "priority": { "narrows_set": 3, "cost_to_answer": 3 },
  "evidence": ["nhtsa_utqg", "tirerack_explainer"]
}
</example>

Judgement that would accompany it:
{
  "claim": "Treadwear grades can be compared across brands as a rough guide but not as a ranking.",
  "confidence": "high",
  "sources_disagree": true,
  "note": "The regulator describes the test as standard. Specialist reviewers say manufacturers calibrate differently. We side with the reviewers because they test the tyres."
}

=== Persona ===
A pharmacist, who thinks in concentration, delivery form and interactions

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

1. **The equivalent of treadwear is the size printed on the cup: "1 cup", "1 tbsp".** It is on every piece and every listing, and it needs two judgements attached. First, it is only as true as the maker's tooling: reviewers who weigh water find most cups within a fraction of a percent, but spoons up to a third off (ATK and Wirecutter, both verbatim below). Second, "cup" and "tablespoon" mean different volumes by country: 236.6 mL (US customary), 240 mL (US nutrition label), 250 mL (metric cup); 14.8, 15 or 20 mL for a tablespoon (NIST, 21 CFR 101.9, Wikipedia, Allrecipes Australia). The ml number printed beside "1 cup" tells you which cup it is: Hudson Essentials marks its 1 cup "240ml"; a NewTalent jug marks "250ml (1-Cup)" and "8 oz" together, which cannot both be true (8 US fl oz is 236.6 mL). **Kept** as `capacity_accuracy` (researched, top N only, from reviewer tests) and as `unit_system` (split 2) with the derived `unit_mismatch_pct`.
2. **The pharmacist notices three things.** (a) Delivery form: the dose depends on the fill method more than the vessel. A level tablespoon of PEG-3350 weighed about 10 g and a heaping one about 15 g (Nichols et al. 2024); a cup of flour weighs 120 g by King Arthur's method and 142 g (5 oz) by ATK's. **Kept** as judgement 2 and in the `flush_rim` facet (a rim you can strike off is what makes "level" repeatable). (b) Concentration: error matters most where the ingredient is potent per gram, which is the spoons (salt, leavening, yeast), and spoons are exactly where accuracy is worst. ATK says to weigh "yeast or xanthan gum". **Kept** in `capacity_accuracy` and judgement 4. (c) Medicine: a kitchen measuring spoon is not a dosing device. The FDA's 2011 guidance wants a calibrated device packaged with every OTC liquid medicine, and the AAP says to use the mL tool that comes with the medicine; yet Spring Chef markets its spoons to "Measure Medicine". **Kept** as judgement 7. **Dropped**: "interactions" as static cling of powder to plastic spoons (ATK, verbatim) as a separate facet; it lives inside `material` because only ATK reports it.

## Sources and what each supports

### Standards bodies and regulators

**NIST SP 811 Appendix B.8, conversion factors (alphabetical)** https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b8
- verbatim table rows: "cup (U.S.) | milliliter (mL) | 2.365 882 E+02"; "tablespoon | milliliter (mL) | 1.478 676 E+01"; "teaspoon | milliliter (mL) | 4.928 922 E+00".
- Supports: US customary cup 236.59 mL, tablespoon 14.79 mL, teaspoon 4.93 mL. Confidence high.

**NIST Metric Kitchen: Culinary Measurement Tips** https://www.nist.gov/pml/owm/metric-si/metric-kitchen/metric-kitchen-culinary-measurement-tips (fetched at https://www.nist.gov/pml/owm/culinary-measurement-tips, which serves the same page)
- verbatim: "Measuring ingredients by mass (weight) using an appropriate kitchen scale is a best practice to improve recipe accuracy, rather than measuring ingredients by volume using fluid or dry measuring cups."
- verbatim: "Using dry measuring cup (designed to measure the volume of dry commodities like sugar) instead of using a fluid measuring cup (designed to measure the volume of liquid commodities like water) is a common way of introducing measurement errors into the cooking process."
- verbatim: "Set the fluid measuring cup on a level surface to accurately read the meniscus. Read the meniscus at eye level." "Strike off excess dry ingredients when using a dry metric measuring cup."
- verbatim: "Remember that “customary” measures, such as teaspoons, cups, pints, quarts, and gallons, mean different things in different countries."
- verbatim: "Some historic recipes may refer to metric "cup and spoon" measures, which are slightly larger than and can often be used interchangeably with the US customary "cup and spoon" measures."
- Supports: scale first; dry and liquid cups are not interchangeable; meniscus and strike-off technique; customary units differ by country. Note NIST calls metric and US cups "often" interchangeable, which is softer than the bakers' line; see judgement 11. Confidence high.

**NIST Metric Kitchen: Cooking Measurement Equivalencies** https://www.nist.gov/pml/owm/metric-si/metric-kitchen/metric-kitchen-cooking-measurement-equivalencies
- verbatim table: "5 mL | 1 teaspoon"; "15 mL | 1 tablespoon (1/2 fluid ounce)"; "240 mL | 1 cup"; footnote "*Approximately".
- Supports: the rounded 240 / 15 / 5 mL equivalents used in US practice. Confidence high.

**NIST Metric Kitchen: Cookware and Tools** https://www.nist.gov/pml/owm/metric-si/metric-kitchen/metric-kitchen-cookware-and-tools
- verbatim: "Measuring ingredients my weight rather than volume is more precise and consistent" (typo in the original).
- verbatim: "Remember, the density of water is about 1 g per mL. Example: If your full baking dish holds 1000 g of water, then your baking dish volume is about 1000 mL or 1 L."
- Supports: the water-weighing check anyone can run on a cup (the method both reviewers use). Confidence high.

**21 CFR 101.9(b)(5)(viii), via Cornell LII** https://www.law.cornell.edu/cfr/text/21/101.9
- verbatim: "For nutrition labeling purposes, a teaspoon means 5 milliliters (mL), a tablespoon means 15 mL, a cup means 240 mL, 1 fl oz means 30 mL, and 1 oz in weight means 28 g."
- Supports: the US "label cup" of 240 mL is a labelling definition, not the customary cup. Confidence high.

**Federal Register 2011-10965, FDA guidance on dosage delivery devices for OTC liquid drugs (5 May 2011)** https://www.federalregister.gov/documents/2011/05/05/2011-10965/guidance-for-industry-on-dosage-delivery-devices-for-orally-ingested-otc-liquid-drug-products
- verbatim: "packaged with dosage delivery devices (e.g., calibrated cups, droppers, syringes, or spoons)"; "FDA is issuing this guidance because of ongoing concerns about potentially serious accidental drug overdoses"; "FDA recommends that dosage delivery devices be included for all orally ingested OTC drug products that are liquid formulations, that they should bear markings that are consistent with the labeled dosage directions, and that they should be labeled in a manner that attempts to ensure that they are used only with the products with which they are included."
- Supports: medicine goes with its own packaged device. The FDA PDF itself (fda.gov/media/79107/download) returned an Akamai "abuse-detection" redirect to curl and 404 to WebFetch; the Federal Register notice of availability is the agency's own summary. Confidence high.

**Federal Register 2012-17366, Indirect Food Additives: Polymers (17 July 2012)** https://www.federalregister.gov/documents/2012/07/17/2012-17366/indirect-food-additives-polymers
- verbatim: "amending the food additive regulations to no longer provide for the use of polycarbonate (PC) resins in infant feeding bottles (baby bottles) and spill-proof cups ... because these uses have been abandoned."
- verbatim: "PC resins may be safely used as articles or components of articles intended for use in producing, manufacturing, packing, processing, preparing, treating, packaging, transporting, or holding food, in accordance with the prescribed conditions of § 177.1580."
- verbatim: "an amendment or revocation based on abandonment is not based on safety".
- Supports: "BPA-free" is not a US legal requirement for kitchen measuring cups; food-contact plastics are regulated under 21 CFR 177. Confidence high.

**American Academy of Pediatrics, HealthyChildren.org: How to Use Liquid Medicines for Children** https://www.healthychildren.org/English/safety-prevention/at-home/medication-safety/Pages/Using-Liquid-Medicines.aspx
- verbatim: "Never use teaspoons, tablespoons or other household spoons to measure medicine."; "Use the dropper, syringe, medicine cup or dosing spoon that comes with the medicine."; "It is easier to measure the right amount with a dosing tool that uses milliliters (mL)."; "never use a kitchen spoon to measure out medicine. This is because kitchen spoons come in lots of different sizes."
- Source line: "American Academy of Pediatrics Council on Quality Improvement and Patient Safety (Copyright © 2021)".
- Note: the AAP's "kitchen spoon" means flatware. Its positive rule (use the mL tool packaged with the medicine) is what applies to kitchen measuring spoons. Confidence high.

### Specialist reviewers who measure

**America's Test Kitchen, The Best Dry Measuring Cups of 2026** https://www.americastestkitchen.com/equipment_reviews/1702-dry-measuring-cups
- verbatim: "We carefully filled each cup to its maximum capacity with water and weighed it on a calibrated lab-quality scale ... Our top performers were either spot-on or off by just a fraction of a percentage point."
- verbatim: "Some cups’ measurements were off by as much as 6 percent when we used them to weigh water. We didn’t recommend these, as this error margin can be enough to affect a recipe’s ratios". Individual sets: "the 1-cup measure was off by nearly 5 percent"; "The largest cup was off by 4 percent"; two sets "the 1-cup measure was off by 6 percent".
- verbatim: "Almost all the stainless-steel sets had etched markings, which we preferred because they remained intact"; "Some plastic sets and one stainless-steel set had measurement markings printed in ink, which came off after very little scrubbing."; the Best Buy set's markings were "seemingly printed on, despite brand assurances that they’re etched".
- verbatim: "Cups with handles that were at least 3 inches long"; "We preferred sets with flat handles that were perfectly flush with the cups’ rims, which allowed us to sweep across the handle and cup"; "We prized cups that weighed 4 ounces or less"; "Most cups made from lighter-weight steel or plastic bent or warped when we repeatedly dug them into densely packed wet sand".
- verbatim: "additional cups in sizes ranging from ⅛ cup to ¾ cup. These extra cups were occasionally helpful ... But some testers felt the added convenience wasn’t worth the extra storage space".
- verbatim: "Some sets were sold with removable connector rings ... Other models had magnets or grooves on their handles that snapped together. All of these methods were equally effective".
- verbatim: "handles topped with or enclosed in soft plastic ... had a tendency to collect dust and particles of flour".
- Disclosure, verbatim: "America’s Test Kitchen and Sur La Table are owned by the same parent company". The winner is a Sur La Table set.
- Supports: accuracy method and spread; etched beats printed; flush rim; extra sizes trade-off; connectors equivalent for cups; handle and weight preferences (notes only). Confidence high.

**America's Test Kitchen, The Best Liquid Measuring Cups of 2026** https://www.americastestkitchen.com/equipment_reviews/2041-liquid-measuring-cups
- verbatim: "we used a lab-grade scale to weigh water at every cup, ounce, and milliliter marking. About half the models we tested were either spot-on or off by just a few drops."; "On the worst models, the 1-cup markings were off by a full tablespoon".
- verbatim: "The benefit of glass is that it’s dishwasher-safe, and many are advertised as microwave-safe too. Glass measuring cups tend to be heavy, though, and they’re fragile."; "Plastic models are lighter ... Even if they are described as dishwasher- and/or microwave-safe, we recommend washing them by hand and avoiding microwaves as the intense heat can cause the plastic to break down and leach."; "silicone isn’t transparent, so it can be harder to read the fill level, and the porous material often retains odors."
- verbatim: "Liquid measuring cups range from 1 to 8 cups in capacity."; "If you’re only getting one liquid measuring cup, get one with a 1-cup capacity. These cups have the most detailed markings ... The 8-cup liquid measuring cup from Pyrex, for example, only has markings for full cups."
- verbatim: "If the top measurement line was too close to the cup’s rim—¼ inch or less—liquid often sloshed ... We were better able to control the flow from cups that offered about 1 inch of clearance".
- verbatim: "Two cups each had two sets of measurement markings: one outside, like usual, and another inside in an upward-facing position ... The markings, which were located on a ridge inside the cup, made it hard to scrape out more-viscous ingredients"; "On one cup with two sets of measurement markings, the markings on the outside of the cup were accurate, but the ones inside were not."
- verbatim: "one plastic measuring cup and the one silicone measuring cup were tinted or had light yellow rings."
- Supports: liquid-cup accuracy spread; material trade-offs; capacity and marking density; read-from-above trade-off; headroom (notes only). Confidence high.

**America's Test Kitchen, The Best Measuring Spoons of 2026** https://www.americastestkitchen.com/equipment_reviews/1920-measuring-spoons
- verbatim: "We think 1 tablespoon, 1 teaspoon, ½ teaspoon, ¼ teaspoon, and ⅛ teaspoon are the most essential measures."; "Two sets lacked the ⅛-teaspoon measure, which we find essential."
- verbatim: "Some 1-tablespoon and 1-teaspoon measures were off by 30 percent or more, translating to 2 to 3 grams of excess or missing ingredients. We downgraded measuring spoons that were off by 10 percent or more."; "the 1-tablespoon measure held about 40 percent more flour than it should have, roughly 3 grams extra"; "The narrower oval 1-teaspoon measure ... about 35 percent, or 1.7 grams, less".
- verbatim: "We often dip measuring spoons directly into small spice jars, which typically have just a 1.5-inch opening. Spoons with narrower bowls, either oval or rectangular in shape, were easier to fit"; Hudson Essentials: "its largest spoon was less than an inch wide".
- verbatim: "Plastic and silicone spoons are nonconductive and hold static electricity—which attracts powdery particles—whereas metal spoons allow the charge to flow, so ingredients don’t cling."
- verbatim: "Our favorite sets were held together with a hinged connector ring."; "A magnetic set also scored poorly because the magnets weren’t strong"; "One of the sets was connected by an open ring, and the gap on the ring made it easy for spoons to slide off".
- verbatim: "if you’re baking and using small amounts of dry ingredients that can make or break a recipe (e.g. yeast or xanthan gum), we recommend a [scale]".
- verbatim (adjustable spoons): "the movable wall ... didn’t form a tight seal with the bottom of the spoon, so ingredients sometimes slipped underneath".
- Supports: spoon accuracy spread; 1/8 tsp essential (ATK side); spice-jar fit; static on plastic; hinged ring preferred. Confidence high.

**America's Test Kitchen, The Best Adjustable Measuring Cups** https://www.americastestkitchen.com/equipment_reviews/2257-adjustable-measuring-cups
- verbatim: "A good adjustable measuring cup can be indispensable when measuring messy, semisolid ingredients such as mayonnaise and honey."; "The cups in our lineup were all accurate"; the winner's "barrel cracked when we dropped it on the floor"; "this model glides most smoothly when a small amount of oil is occasionally applied to the plunger tip"; "This cup isn't intended to replace dry or liquid measuring cups".
- Supports: the adjustable family, its job and its fragility. Confidence high (one source; no number used).

**America's Test Kitchen, Dry versus Liquid Measuring Cups (how-to)** https://www.americastestkitchen.com/how_tos/5450-dry-versus-liquid-measuring-cups
- verbatim: "we asked 18 people ... to measure 1 cup of all-purpose flour ... With the dry measuring cup, the measurements were off by as much as 13 percent ... Measuring flour in a liquid measuring cup, where it’s impossible to level off any excess, drove that variance all the way up to 26 percent."; water: "The dry cup varied by 23 percent, while the liquid cup varied by only 10 percent."
- Supports: wrong-type error is larger than tool error (numbers in notes only; one source). Confidence high.

**America's Test Kitchen, Measuring Flour (how-to)** https://www.americastestkitchen.com/how_tos/6573-measuring-flour
- verbatim: "Our recipes list 1 cup of flour as weighing 5 ounces"; "we’ve found that there can be up to a 20 percent difference in the weight of a cup of flour"; "do not be tempted to transfer the weight conventions from one recipe to another."
- Supports: ATK's 5 oz (142 g) flour cup, by dip-and-sweep. Confidence high.

**America's Test Kitchen, Dry vs. Liquid Measuring Cups for Thick Wet Ingredients (how-to)** https://www.americastestkitchen.com/how_tos/6568-dry-vs-liquid-measuring-cups-for-thick-wet-ingredients
- verbatim: "The fluctuation in measurements using the Adjust-A-Cup varied by only 2 grams, but the fluctuation using the dry and liquid measures was as much as 14 grams."; "If you don’t own one, a dry measuring cup is the next most consistent tool."
- Supports: adjustable cup for semisolids (number in notes only). Confidence high.

**America's Test Kitchen, Do You Really Have to Use Different Measuring Cups (article)** https://www.americastestkitchen.com/articles/5516-do-you-really-have-to-use-different-measuring-cups-for-liquid-and-dry-ingredients
- verbatim: "Liquid measuring cups have extra space at the top to prevent sloshing or spilling liquids."; "Because the measurement markings are down inside the cup, you can’t sweep across the top".
- Supports: the mechanism of the two cup types. Confidence high.

**Wirecutter, The Best Measuring Spoons (updated 28 October 2025), via Internet Archive snapshot of July 2026** https://www.nytimes.com/wirecutter/reviews/best-measuring-spoons/ (live page blocks curl and WebFetch; read at https://web.archive.org/web/2026/https://www.nytimes.com/wirecutter/reviews/best-measuring-spoons/)
- verbatim: "the tablespoons in a couple of other sets we tested were off by as much as 5 grams — the equivalent of a teaspoon (or a third of a tablespoon!)."
- verbatim: "We filled the spoons as precisely as possible, repeated the measurement 10 times for each spoon ... compared that average to the standard weight (in milliliters), defined by the National Institute of Standards and Technology".
- verbatim: "We preferred spoons with oval or rectangular bowls, which are easier to fit into the narrow mouth of a spice jar."
- verbatim: "A ⅛ teaspoon can offer extra accuracy, but for amounts that small, it usually won’t hurt to just eyeball half of the ¼ teaspoon."
- verbatim: "plastic spoons usually have painted-on measurement markings that can wear off over time, whereas most metal spoons have stamped or etched measurements."; "at least one of her sets of plastic spoons seems to have warped in the dishwasher."
- verbatim: "Magnetic fasteners are a much more convenient way than a ring to keep spoons together"; "Some Amazon reviewers complain that the magnets have fallen out of some of their spoons."
- verbatim (double-ended Prepworks): "the round ends were about as accurate as the Sur La Table spoons ... the oval ends were less impressive. Some of the oval spoons measured multiple grams off the mark, with the tablespoon falling short by about a quarter teaspoon."
- verbatim: "We also like that the bowls aligned with the handles perfectly (it’s all one piece of metal), so we could easily sweep a knife across to level ingredients."
- Supports: spoon error up to a third (second source to ATK); spice-jar fit; 1/8 tsp a bonus (disagrees with ATK); printed versus stamped; magnets preferred (disagrees with ATK spoons); double-ended trade-off; flush handle. Confidence high.

**Could not reach**: Wirecutter's measuring cups review (curl blocked, WebFetch refused nytimes.com, no Wayback snapshot of that URL); Serious Eats (people.inc block page to curl, no Wayback snapshot); OXO.com (Cloudflare challenge); FDA PDFs (Akamai block). The dry-cup and liquid-cup accuracy figures therefore rest on ATK alone and are kept in these notes, not in the JSON.

### Academic

**Nichols KR et al., Measurement of Polyethylene Glycol 3350 With Standard Household Measuring Devices, J Pediatr Pharmacol Ther 2024;29(3):286, doi 10.5863/1551-6776-29.3.286** (abstract read via Europe PMC REST API, https://www.ebi.ac.uk/europepmc/webservices/rest/search?query=EXT_ID:38863850&format=json&resultType=core)
- verbatim: "A leveled teaspoonful of PEG-3350 weighed ∼3.3 grams and an unleveled teaspoonful weighed ∼3.7 grams. A leveled, unleveled, and heaping tablespoon of PEG-3350 weighed about 10, 11, and 15 grams, respectively."; "Using 1 dedicated measurement device and leveling the dose may improve consistency".
- Supports: fill method moves a spoonful by half (seed item 2a; one source, notes only). Confidence medium (abstract only).

### Primers

**Wikipedia, Cup (unit)** (raw wikitext) https://en.wikipedia.org/wiki/Cup_(unit)
- verbatim: "Australia, Canada, New Zealand, and some other members of the Commonwealth of Nations ... employ a "metric cup" of 250 millilitres."; "The cup currently used in the United States for nutrition labelling is defined in United States law as 240 ml."; US customary "236.5882365 | millilitres exactly".
- Supports: 250 / 240 / 236.6 mL cups. Confidence high.

**Wikipedia, Tablespoon** (raw wikitext) https://en.wikipedia.org/wiki/Tablespoon
- verbatim: "The international metric tablespoon, used in the UK and US is exactly 15 mL."; "An Australian metric tablespoon is 20 mL."; "In Australia, one tablespoon equals four teaspoons."; "This definition was promulgated by the Metric Conversion Board in the 1970s".
- Supports: the Australian 20 mL tablespoon. Confidence high.

**Wikipedia, Measuring spoon** (raw wikitext) https://en.wikipedia.org/wiki/Measuring_spoon
- verbatim: "The volume of a traditional US teaspoon is 4.9 ml and that of a tablespoon is 14.8 ml, only slightly less than standard metric measuring spoons."; Japan: "A large spoon is 15 milliliters, and a small spoon is 5 milliliters."
- Supports: US spoons versus metric; Japan uses the 15/5 mL pair. Confidence high.

**Wikipedia, Measuring cup** (raw wikitext) https://en.wikipedia.org/wiki/Measuring_cup
- verbatim: "Dry measure cups are distinguished from liquid measure cups in that they are meant to be filled to the top so that excess may be scraped off"; "Liquid measure cups tend to be microwave safe for heating and clear to more easily judge the meniscus."; "1 cup of all-purpose flour sifted into a cup and leveled weighs about 100 grams (3½ oz), whereas 1 cup of all-purpose flour scooped from its container and leveled weighs about 140 grams (5 oz)." (citing Child and Beck).
- Supports: mechanism; flour weight depends on method (second source beside ATK's 5 oz). Confidence medium-high.

**Wikipedia, Pyrex** (raw wikitext) https://en.wikipedia.org/wiki/Pyrex
- verbatim: "Starting around the 1950s, Corning began making Pyrex out of thermally tempered soda–lime glass, instead of borosilicate glass"; "soda–lime glass has higher mechanical strength than borosilicate—it has around twice the strength ... However, its thermal shock resistance is lower than borosilicate's".
- Supports: tempered versus borosilicate glass trade-off. Confidence medium-high.

**Allrecipes Australia, Standard Australian Cooking Measurements (Internet Archive snapshot of 6 October 2014)** https://web.archive.org/web/20141006142213/http://allrecipes.com.au/how-to/17/standard-australian-cooking-measurements.aspx
- verbatim: "The standard cup used in Australian and NZ Cooking is 250ml, the standard metric cup."; table "1 tablespoon | 20ml"; "1 teaspoon | 5ml".
- Also verbatim: "The U.S. cup is 240 ml"; "The U.K cup is 225ml". The 225 mL UK cup disagrees with Wikipedia ("there is no imperial unit–based culinary measuring cup") and is not used.
- Supports: second source for 250 mL cup and 20 mL tablespoon. Confidence high for those two.

**King Arthur Baking, Ingredient Weight Chart** https://www.kingarthurbaking.com/learn/ingredient-weight-chart
- verbatim: "For best results, we recommend weighing your ingredients with a digital scale. A cup of all-purpose flour weighs 120 grams or 4 1/4 ounces."
- Supports: King Arthur's 120 g flour cup (against ATK's 142 g). Confidence high.

### Brand spec sheets (fields brands list, in their order)

Read as Shopify product JSON (`/products.json`), which carries the product page body text.

**Hudson Essentials, Stainless Steel Measuring Cups Set, 6 pcs** https://hudsonessentials.com/products/hudson-essentials-stainless-steel-measuring-cups-set-6-piece-stackable-set-with-spout
- verbatim order: "Heavy Duty Construction: 18/8 stainless steel with heavy duty single-piece construction & (no-welds, no-rivets, or soldered handles)"; "stamped measurements on cups"; "partial measurements are stamped & visible on the outside and inside of cups"; "Precise Measurements: both metric and standard"; "stackable, nesting design"; "dishwasher-safe"; sizes "1 cup (240ml) - 3/4 cup (180ml) - 2/3 cup (160ml) | 1/2 cup (120ml) - 1/3 cup (80ml) - 1/4 cup (60ml)".
- Supports: brands mark the 240 mL label cup in ml; field list. Confidence high.

**Hudson Essentials, Measuring Spoons Set of 7** https://hudsonessentials.com/products/hudson-essentials-stainless-steel-measuring-spoons-set-for-dry-or-liquid-fits-in-spice-jars-set-of-7
- verbatim: "includes 1/8 tsp, 1/4 tsp ,1/2 tsp, 3/4 tsp, 1 tsp, 1/2 tbsp, 1 tbsp"; "Rectangular and narrow measuring spoons fit easily in most spice containers and stamped measurements on every handle"; "easy-to-read measurements in both metric and standard".

**Cuisipro, Stainless Steel Measuring Cups and Spoon Set** https://cuisipro.com/products/cuisipro-silver-stainless-steel-stainless-steel-measuring-cups-and-spoon-set-two-sets-747143
- verbatim: "oval to reach into the smallest spice jar"; "Gently curved handles allow the spoons to sit securely on the counter without tipping over"; "1/8, ¼, ½, 1tsp, 1tbsp/0.625, 1.25, 2.5, 5, 15ml"; "held together with a simple storage clip".
- Supports: spoons marked on the 5/15 mL metric basis; spice-jar claim; clip connector. Confidence high.

**Spring Chef, Stainless Steel Measuring Cups, Set of 7** https://springchef.com/products/spring-chef-measuring-cups-stainless-steel-set-of-7
- verbatim: "Measurement values are engraved on the handles in both U.S. and Metric units, so they won't rub or fade away."; "1/8 cup, 1/4 cup, 1/3 cup, 1/2 cup, 2/3 cup, 3/4 cup, and 1 cup plus a ring".

**Spring Chef, Magnetic Measuring Spoons Set of 8** https://springchef.com/products/spring-chef-magnetic-measuring-spoons-set-dual-sided-stainless-steel-fits-in-spice-jars-set-of-8
- verbatim: "The oval end is narrow to fit into most spice jars."; "The round end is perfect for liquids."; heading "Complete, Accurate Set To Measure Spices, Medicine And More".
- Supports: brands market measuring spoons for medicine (judgement 7). Confidence high.

**Pyrex, Product Warranties, Safety and Usage** https://pyrexhome.com/pages/product-warranties-safety-and-usage
- verbatim: "AVOID sudden temperature changes, to avoid the risk of breakage. DO NOT add liquid to hot glassware"; "DO NOT use on or under a flame or other direct heat source, including on a stovetop, under a broiler, on a grill"; "Can be used for cooking, baking, warming, and reheating food in microwave ovens".
- Supports: glass liquid cups go in the microwave, never on the stove; thermal-shock warnings. Confidence high.

### Retail listings (what data is present)

Amazon pages fetched with curl (US store, rendered for an India delivery address; prices ignored).

- **Search "measuring cups", "measuring spoons", "liquid measuring cup"** (https://www.amazon.com/s?k=measuring+cups and so on). Titles carry: material ("Stainless Steel", "Glass", "Plastic", "Borosilicate"), piece count, sizes for liquid cups ("1, 2 & 4 Cup"), "Magnetic", "Fits in Spice Jars", "Dishwasher Safe", "BPA-Free", "Microwave Safe", "Read Measurements from Above", "Laser-Etched", and ml capacities ("250ml (1-Cup), 500ml (2-Cup)"; "Etched Glass Measuring Cup with Handle, 500 ml, 2 Cup").
- **OXO Good Grips Angled Measuring Cups, 3-Piece (B0000CCY1Y)**: bullets "read measurements from above", "1-Cup, 2-Cup, and 4-Cup ... marked with cups, ounces, and milliliters", "measurement markings ... will stay put, resisting wear", "BPA-free ... Tritan Renew". Table: "Material Type : Tritan", "Capacity : 32 fluid ounces".
- **OXO Good Grips 4 Piece Stainless Steel Measuring Cups with Magnetic Snaps (B00L9WJZW2)**: "Set includes: 1/4 cup, 1/3 cup, 1/2 cup and 1 cup"; "Permanent, etched measurement markings"; "Designed to make leveling ingredients easy"; "Item Weight : 12.64 ounces"; "Material Type : Stainless Steel".
- **Pyrex Essentials 3-Pack (B00M2J7PCI)**: "High-quality tempered glass is dishwasher, freezer, microwave and preheated oven safe."; "bold and large measurements in both ounces and milliliters"; "Material Type : Glass".
- **Hudson Essentials cups (B00XWDLBKK) and spoons (B07V2KT8GR)**: sizes in bullets; "stamped measurements"; "Square and narrow measuring spoons fit easily in most spice containers"; table "Capacity : 1 ml" on a 7-spoon set.
- **Spring Chef Magnetic Spoons (B07DJ53GZQ)**: "U.S. and metric measurement markings"; "The oval end is narrow to fit into most spice jars"; "Measure Medicine And More"; "Capacity : 15 ml".
- **Amazon Basics 10-piece plastic set (B0DZXCBMBV)**: sizes in "Included Components"; "food-grade plastic"; "Loop design"; "Capacity : 2366 ml" for a set whose biggest cup is 1 cup.
- **NewTalent 3-pack plastic jugs (B0HBQCCKQ4)**: "Large（1000ml，4 cups,32 oz), medium (500ml,2 cups,16 oz), small(250ml,1 cups,8 oz)"; "withstand temperatures from 0°C to 80°C (32°F to 176°F)"; "Perfect for accurate liquid measuring cups or dry measuring cups".
- What this tells the scraper: material, piece list and liquid capacity are on every page fetched (8 of 8). Units are in bullets on 6 of 8. Marking type is in bullets on 4 of 8 (OXO steel, Hudson both, Spring Chef) and only when it is etched, stamped or engraved; printed markings are never announced. Spice-jar fit appears as a title or bullet claim on spoon sets that have it. Amazon's "Capacity" row is noise for sets (2366 ml, 1 ml, 15 ml, "1 cups").

## Data points used for derived metrics

| Measure | mL | Source |
|---|---|---|
| US customary cup | 236.59 | NIST B.8; Wikipedia Cup |
| US nutrition-label cup | 240 | 21 CFR 101.9; NIST equivalencies; Wikipedia Cup |
| Metric cup (AU, CA, NZ) | 250 | Wikipedia Cup; Allrecipes AU |
| US customary tablespoon | 14.79 | NIST B.8; Wikipedia Measuring spoon (14.8) |
| Metric and label tablespoon | 15 | 21 CFR 101.9; Wikipedia Tablespoon; Cuisipro |
| Australian tablespoon | 20 | Wikipedia Tablespoon; Allrecipes AU |
| US customary teaspoon | 4.93 | NIST B.8; Wikipedia Measuring spoon (4.9) |
| Metric and label teaspoon | 5 | 21 CFR 101.9; Allrecipes AU; Cuisipro |

`unit_mismatch_pct` worked examples (arithmetic on the rows above): metric cup in a US recipe 250 / 236.59 − 1 = +5.7%; US-label cup marking 240 / 236.59 − 1 = +1.4%; Australian tablespoon in a US or UK recipe 20 / 14.79 − 1 = +35% (20 / 15 − 1 = +33% against the metric tablespoon); US cup in an Australian recipe 236.59 / 250 − 1 = −5.4%.

`capacity_error_pct` inputs: water at about 1 g per mL (NIST Cookware and Tools), the nominal mL from the table above for the tool's own system, and the mean of several weighings (ATK: three for cups, five for spoons; Wirecutter: ten for spoons).

Spoon error, two sources: ATK "off by 30 percent or more" (water) and 40 percent (flour) on 1 tbsp; Wirecutter "off by as much as 5 grams" on a tablespoon, which is 5 / 14.79 = 34%. The JSON says "about a third".

Flour cup, two numbers each from its owner: King Arthur 120 g; ATK 5 oz = 141.7 g; Wikipedia (Child and Beck) 100 g sifted, 140 g scooped. The scooped figure corroborates ATK's.

## Against the existing add-on schema

Add-on read: `supplies/kitchen/add-ons/measuring_cup.jsonc` (one field, `units_marked` enum [us_customary, metric, both, other, unspecified]), composed with `kitchen-standard.jsonc` (`materials[]` with roles, `max_heat_temp`, `dishwasher_safe`) and `supplies-standard.jsonc` (`high_signal_markdown`, `sizing` with `size_stated`, `pack_count`, `pieces_included_text`, `dimensions`, `capacity`, `weight`, `measurements_text`). Kinds follow AGENTS.md: vertical anchor, quality currency, hard filter, sizing, or markdown.

| Facet in this register | In schema today | Proposed kind under AGENTS.md | Note |
|---|---|---|---|
| measure_type | no | vertical anchor (new enum) | Proposal: `measure_type` [dry_cup_set, liquid_cup, spoon_set, cups_and_spoons_set, adjustable_cup]. The leaf mixes four tools that experts review separately; every other facet depends on it. Title text settles it. |
| unit_system | yes, `units_marked` | hard filter (existing, widen) | Add `australian_metric` (20 mL tablespoon). Define `both` as two scales on one liquid cup, or two labels on one dry piece; for a dry piece with two labels, record which mL the "1 cup" carries in markdown (240 versus 250 decides the system). |
| spice_jar_fit | no | high_signal_markdown | Bowl shape is a photo fact (rule 1); the "fits in spice jars" claim is text and worth one markdown phrase. Not a column. |
| capacity_accuracy | no | outside schema (researched) | Only reviewer water-weighing produces it. Top N from ATK and Wirecutter tests; lazy-load the rest. |
| marking_durability | no | high_signal_markdown | Brands state it only when it is etched, stamped or engraved, so absence is weak evidence of print. Belongs in the build-quality line the markdown comment already asks for. |
| flush_rim | no | image pipeline | A photo answers it (rule 1). No field. |
| sizes_included | yes, `sizing.pieces_included_text`, `sizing.pack_count` | sizing (existing) | Keep. The odd sizes (1/8 tsp, 2/3 and 3/4 cup) are what a buyer filters on; they come out of this text at query time. |
| material | yes, `materials[]` | vertical anchor (existing) | Fits as is: stainless_steel, tritan_copolyester, plastic, glass, tempered_glass, borosilicate_glass, silicone. Record tempered versus borosilicate when stated. |
| microwave_safe | no; `max_heat_temp` is near it | hard filter (proposal, kitchen-standard level) | Proposal: `microwave_safe` enum [yes, no, unspecified] in kitchen-standard, since bowls and containers share it. For liquid cups it is the glass-versus-plastic decision. Note ATK advises against microwaving plastic even when labelled safe. |
| read_from_above | no | high_signal_markdown | OXO's angled insert is named in titles; a photo partly shows it. One markdown phrase. |
| liquid_capacity | partly, `sizing.capacity` | sizing (existing) | `capacity.unit` has no `cup`; store as fl_oz or ml with the cup size in `size_stated`. Do not ingest Amazon's "Capacity" row for sets (2366 ml for a 1-cup set). |
| connector | no | high_signal_markdown | Ring, clip or magnets are in titles and photos; sources disagree on which is better. Not a column. |
| unit_mismatch_pct | no (derived) | derived at query time | From `units_marked` plus the recipe's system. Not stored. |

Fields this register considers noise for measuring cups and spoons: `sizing.dimensions` (a photo shows size, and nobody filters a cup set by its footprint); `sizing.weight` for a whole set (ATK cares about the weight of one cup, which listings never give; set weight tells you nothing); `max_heat_temp` for dry cups and spoons (only relevant to plastic liquid cups, where `microwave_safe` is the real question); `dishwasher_safe` is near-universal on metal and glass but worth keeping for plastic sets, where Wirecutter reports warping.

## Branches

Each leg names the register sentence or the notes sentence it follows from. Split 1's `when` labels are its values list; the question's answers map one to one ("dry cups" to dry cup set, "liquid jug" to liquid measuring cup, "spoons" to spoon set, "plunger cup" to adjustable plunger cup), and a buyer who answers both "dry cups" and "spoons" maps to cups and spoons set, as the split rationale states.

**Split 1, measure_type ("Dry cups, liquid jug, spoons, or plunger cup for sticky stuff?")**

- when `dry cup set`: then_ask material, flush_rim, marking_durability, sizes_included; makes_irrelevant read_from_above, liquid_capacity, spice_jar_fit.
  - material: this trace line was missing before 2026-10-25 (blind check finding 4); `category.json` already carried `material` in `then_ask` here. It is the one askable facet that splits dry sets hard (narrows_set 4, steel against plastic); "light steel and plastic bent in ATK's tests" (material, why experts care).
  - flush_rim: "Dry cups and spoons are filled past the brim and struck level, so a handle flush with the rim decides whether a knife sweeps clean" (flush_rim, why experts care).
  - marking_durability: "Printed ink wore off after little scrubbing in ATK's test ... etched, stamped and engraved marks on steel did not" (marking_durability, why experts care).
  - sizes_included: "Extra 2/3 and 3/4 cups save a second scoop" (sizes_included, why experts care).
  - read_from_above and liquid_capacity irrelevant: both definitions say "a liquid measuring cup".
  - spice_jar_fit irrelevant: definition says "the set's tablespoon".
- when `liquid measuring cup`: then_ask material, microwave_safe, liquid_capacity, read_from_above; makes_irrelevant flush_rim, spice_jar_fit, connector.
  - material and microwave_safe: "For jugs, glass microwaves and resists stains but is heavy and breaks; plastic is light but ATK keeps it out of the microwave" (material, why experts care) and microwave_safe's definition ("Whether the maker rates a liquid cup for heating food in the microwave").
  - liquid_capacity: "A 1-cup jug has the finest markings, down to quarter and third cups; bigger jugs mark fewer steps" (liquid_capacity, why experts care).
  - read_from_above: its definition is liquid-only.
  - flush_rim irrelevant: chain of two stated facts. (1) flush_rim matters only when you strike off past the brim (its why). (2) "A liquid cup is a clear jug with its marks below the rim, so you never fill to the brim" (mechanism). So a level-off edge has no job on a jug.
  - connector irrelevant: its definition is "how a nesting set of dry cups or spoons is held together".
- when `spoon set`: then_ask spice_jar_fit, sizes_included, material, marking_durability, connector; makes_irrelevant read_from_above, liquid_capacity. flush_rim is not made irrelevant, because spoons are struck level too (flush_rim definition names spoons).
  - spice_jar_fit: its definition (spoons).
  - sizes_included: "Spoon sets differ on the 1/8 teaspoon, which ATK calls essential and Wirecutter a bonus" (sizes_included, why).
  - material: "plastic and silicone hold static, so powders cling" (material, why).
  - marking_durability: trace line was missing before 2026-10-25 (blind check finding 15); `category.json` already carried it. The spoon evidence is the strongest of any branch for this facet: Wirecutter's "plastic spoons usually have painted-on measurement markings that can wear off over time" is about spoons specifically.
  - connector: trace line was missing before 2026-10-25 (blind check finding 15); `category.json` already carried it. Spoons are the contested case both reviewers argue about: ATK favours a hinged ring and faults weak magnets, Wirecutter prefers magnets (judgement 6).
  - read_from_above, liquid_capacity: liquid-only definitions.
- when `cups and spoons set`: then_ask material, spice_jar_fit, flush_rim, marking_durability; makes_irrelevant read_from_above, liquid_capacity.
  - material: trace line was missing before 2026-10-25, same basis as the dry leg above; `category.json` already carried it.
  - spice_jar_fit, flush_rim, marking_durability: same sentences as the dry and spoon legs above.
  - read_from_above and liquid_capacity irrelevant: `measure_type`'s own definition states the assumption directly — "a set of dry cups and spoons with no jug" — so a combined set by definition carries no liquid-cup scale to read from above or size. This sentence resolves blind check finding 10, which flagged the assumption as unstated; `category.json`'s definition already stated it.
- when `adjustable plunger cup`: makes_irrelevant spice_jar_fit, connector, read_from_above. `category.json` does not make flush_rim irrelevant here (blind check finding 5, already resolved before this check): the register's own mechanism has the plunger cup "fill, level, push it out", which is a level-off step like a dry cup's, so the facet's irrelevance could not rest only on its definition naming "a dry cup's or spoon's handle". It is not added to then_ask either, since the facet is still defined around a handle flush with a cup's rim, which a plunger barrel does not have; the branch leaves it unclaimed rather than asserting either way.
  - spice_jar_fit, connector, read_from_above: definitions are "the set's tablespoon", "a nesting set of dry cups or spoons" and "a liquid measuring cup" respectively.

**Split 2, unit_system**: no branches. The answer does not change which other facet is asked; it only filters on its own values. The rationale in `category.json` already reads "NIST says the cups can often be swapped; Allrecipes Australia says convert for critical baking" rather than asserting the split "skews every recipe"; this note is stale no longer, since that overstated phrasing would have disagreed with NIST's own "can often be used interchangeably" (judgement 11, blind check finding 7).

`spice_jar_fit` is not a top-level split (blind check finding 6, already resolved before this check; `category.json`'s `splits` array holds only `measure_type` and `unit_system`): it changed nothing for three of the five `measure_type` values and is covered by its place in the `spoon set` and `cups and spoons set` branches above. It keeps its `cold_start_question` for use there, and its `source`/`backfill_cost` match `flush_rim`'s handling (researched, medium; claim from listing text, else the image pipeline for bowl shape).

Dropped legs: "metric → makes sizes_included irrelevant" (no source says metric sets differ in fractions); "spoon set → then_ask capacity_accuracy" (accuracy is not askable; it is an annotation on every branch).

## Numbers in category.json and the sources behind each

| Number | Where | Sources |
|---|---|---|
| 236.6 mL US cup | unit_system, mechanism, vocabulary, metric | NIST B.8; Wikipedia Cup |
| 240 mL label cup | unit_system, vocabulary, judgement 11 | 21 CFR 101.9; NIST equivalencies; Wikipedia Cup |
| 250 mL metric cup | unit_system, vocabulary, judgement 11 | Wikipedia Cup; Allrecipes AU |
| 5.7 percent | unit_system, split 2 rationale, metric | arithmetic on the two rows above |
| 14.8 / 15 / 20 mL tablespoon | unit_system, vocabulary | NIST B.8 and Wikipedia Measuring spoon; 21 CFR and Wikipedia Tablespoon; Wikipedia Tablespoon and Allrecipes AU |
| a third (spoon error) | capacity_accuracy, judgement 4 | ATK spoons (30 to 40 percent); Wirecutter (5 g on a tablespoon) |
| a third (Australian tablespoon) | unit_system, split 2 rationale | arithmetic, 20 / 15 |
| 120 g and 142 g flour cup | judgement 2 | King Arthur; ATK (5 oz), with Wikipedia's 140 g scooped as corroboration |
| 1 g per mL water | capacity_accuracy, metric | NIST Cookware and Tools; Wirecutter ("1 mL = 1 gram") |

Kept out of the JSON because one source carries them: ATK's 6 percent worst dry cup and "a full tablespoon" worst liquid cup; ATK's 10 percent downgrade line for spoons; ATK's 1.5-inch spice-jar mouth; the 1-inch headroom; 3-inch handles; 4-ounce cups; the 13, 26, 23 and 10 percent dry-versus-liquid variances; the 2 g versus 14 g adjustable-cup spread; the PEG 10 g versus 15 g tablespoon.

## Trimmed detail

Facts cut from capped fields, kept here.

- **summary**: the scale caveat ("a scale beats all of them for baking", NIST and ATK) moved to judgement 1.
- **mechanism.plain**: "Liquid cups need about an inch between the top mark and the rim to pour cleanly" (ATK only; trimmed for the source rule, not length). "The adjustable cup's winner cracked when dropped" (in the trade-off and ATK reference).
- **capacity_accuracy.why_experts_care**: ATK's worst dry cups 4 to 6 percent off and its worst liquid cups a tablespoon off at 1 cup; ATK downgrades spoons beyond 10 percent. One source each.
- **spice_jar_fit.why_experts_care**: ATK measures a typical spice-jar mouth at 1.5 inches; Hudson's largest spoon is under an inch wide (ATK). One source.
- **material.why_experts_care**: ATK found light steel and plastic cups bent when dragged through wet sand; plastic and silicone liquid cups stained yellow after turmeric marinara; silicone retained odours.
- **flush_rim.why_experts_care**: soft plastic handle tops collect flour (ATK); ridged spoon handles trap ingredients (ATK); ATK prefers handles at least 3 inches long and cups of 4 oz or less.
- **liquid_capacity.why_experts_care**: Pyrex's 8-cup jug marks only whole cups (ATK).

## Script problems

- **Stale: `vertical: supplies` schema failure, corrected 2026-10-25.** This note previously said `validate.py` rejected `"vertical": "supplies"` because the schema enum was [formulations, machines, fashion, furnishing, other]. The skill's `schema.json` enum now reads [formulations, machines, fashion, furnishing, supplies, other], `supplies` is a valid value, and `validate.py` reports zero problems on the real file. The same stale note was flagged in `cocktail-mixing-tools/notes.md`, which is outside this register's scope and not edited here.
- **Viewer check.** `tests/viewer-browser.py` on a scratch copy of viewer.html passed 6 of 6 (1380, 768 and 390 px, light and dark). The first screen at 1380 light shows the tree and the full priority map; the five-way first split makes the vertical tree wider than its panel, so it scrolls sideways, as the viewer design intends.
- **Web search budget.** The session's 200 WebSearch calls were used up before this run started. Sources were found by direct URLs, site sitemaps (ATK), Shopify `products.json` endpoints, the Europe PMC API and the Federal Register API.

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

A blind checker graded this register in `check.md`; this log works through its 16 numbered "Required fixes" plus the Part D hint findings it folds into them. Most of the fixes were already present in `category.json` when this pass started; where that is so, the line says so rather than claiming a new edit, and the stale documentation that still described the old, wrong state (in `notes.md`) is what got corrected.

1. **unit_system mapping for dry pieces.** Already satisfied in `category.json`: the rationale reads "Only jugs serve both", `why_experts_care` already maps "US recipes to US tools, metric to metric, Australian to Australian metric", `source` is `researched` and `backfill_cost` is `medium`. No JSON edit needed.
2. **Cup accuracy wording (capacity_accuracy, judgement 4).** Already satisfied: `why_experts_care` attributes the finding to "ATK, the only cup tester we could read" and no longer claims most cups are within a fraction of a percent; judgement 4's note already reads "though most were acceptable" rather than "Both found the best sets dead on". No JSON edit needed.
3. **The 240 mL tool case.** Already satisfied: `unit_mismatch_pct.why` now says "Treat a tool printed '240 mL' as a US cup" instead of charging it a separate 1.4 percent mismatch, and the "Which cup is a cup" grammar says a 240 mL tool "lands beside the US line". No source for a vessel built to 240 mL was found or needed, since the claim was removed rather than sourced. No JSON edit needed.
4. **Material on the dry and combined branches.** Already satisfied: `material` is already first in `then_ask` for both `dry cup set` and `cups and spoons set`. The `notes.md` "Branches" trace was stale (it omitted the `material` leg on both); fixed in this pass.
5. **Plunger-cup flush_rim leg.** Already satisfied: `adjustable plunger cup`'s `makes_irrelevant` no longer lists `flush_rim`. The `notes.md` trace still argued for the old leg; fixed in this pass.
6. **spice_jar_fit source.** Already satisfied: `spice_jar_fit.source` is `researched` at `medium` cost, with a `backfill_note` that matches `flush_rim`'s handling ("send the photo to the image pipeline... as for flush_rim"), and it is no longer in the top-level `splits` array (only `measure_type` and `unit_system` are). The `notes.md` trace still called it "split 3"; fixed in this pass.
7. **Split 2 (unit_system) as a top-level split.** Already satisfied via the softer-rationale option: the rationale no longer claims the mismatch "skews every recipe" and instead states NIST's and Allrecipes Australia's disagreement plainly. Kept as a top-level split rather than demoted to a branch, since in a single-region store it still answers a real "which system" question before the finer facets. `notes.md` updated to stop describing the old rationale.
8. **A second source for 250 mL.** Already satisfied: `hc_food_guide` (Health Canada's food guide, "250 mL (1 cup)") is in the register's references and in `unit_system`'s and the visual grammar's evidence lists, independent of Wikipedia's own citation of Allrecipes Australia. No new fetch needed.
9. **Scope orphans with siblings.** Fixed in this pass, in `notes.md` only (no sibling file touched, per instructions). Lone measuring and coffee scoops: the "Not in scope" bullet said they go to `manual-coffee-tea-prep`, contradicting `measure_type`'s own definition ("a lone measuring or coffee scoop counts with spoons") and `scoops/notes.md`, which sends them back here; reworded to claim them here, matching the facet definition that was already correct. Bowl sets padded with measuring cups: already claimed by `measure_type`'s `backfill_note` ("classed by its measures, usually cups and spoons set"); added a line in "Not in scope" to state that explicitly, since the register claimed it but the scope note didn't say so.
10. **Combined-set legs (read_from_above, liquid_capacity irrelevant).** Already satisfied: `measure_type`'s definition states the assumption directly, "a set of dry cups and spoons with no jug". The `notes.md` trace didn't point to that sentence; fixed in this pass.
11. **Glass claim in material.** Already satisfied: `why_experts_care` says "Most US Pyrex is tempered soda-lime glass", scoped to Pyrex, not "most US glass jugs". No JSON edit needed.
12. **ATK flour wording.** Already satisfied: the `atk_flour` reference's `supports` line reads "ATK's 5 oz flour cup is the average of its volunteers' cups" (not "by dip and sweep"), and judgement 2's note no longer claims King Arthur's figure comes "from different filling methods". No JSON edit needed.
13. **sizes_included evidence.** Already satisfied: the unsourced "Four cups and four or five spoons cover most recipes" claim is not in `why_experts_care`; the field states only the sourced 2/3-and-3/4-cup trade-off and the ATK/Wirecutter essential-spoons disagreement. No JSON edit needed.
14. **Hints (marking_durability, liquid_capacity, sizes_included).** Already satisfied: `marking_durability`'s hint already says only etched/stamped/engraved/laser-etched words count and that durability wording like "stay put, resisting wear" is not evidence; `liquid_capacity`'s hint already points at "Title or bullets" first; `sizes_included.backfill_cost` is already `low`, not `none`. No JSON edit needed.
15. **Spoon branch (marking_durability, connector).** Already satisfied: both are already in `spoon set`'s `then_ask`. The `notes.md` trace omitted both legs; fixed in this pass.
16. **Stale note on the schema enum.** Fixed in this pass: the "Script problems" note in `notes.md` still said `validate.py` rejected `vertical: supplies`; the skill's `schema.json` enum already includes `supplies` and the file validates clean, so the note is rewritten to say so and marked stale-corrected.

Not accepted: none of the 16 findings are rejected. Two of the checker's own observations do not hold against the file as it stands and are noted, not logged as corrections, because nothing needed fixing: the Part B narrowing discussion (d) and Part C "Missing legs" list both describe `material` as absent from the dry and combined branches and the spoon branch as missing `marking_durability`/`connector`; `category.json`'s `then_ask` arrays already carried all of these before this pass (see 4 and 15 above), so the checker appears to have graded against an earlier draft of the JSON that predates this file.