# Foot files & pumice stones: research notes

Date: 2026-10-25. Scope: abrasive tools and materials bought to remove calluses and thickened, hardened skin from feet — natural pumice stones, glass and ceramic foot files, stainless-steel rasp/grater-style foot files (PedEgg-style), emery-paper files, and battery-powered motorized callus removers with replaceable roller heads (Amope Pedi Perfect-style). In the Zamana taxonomy this is `supplies > personal > foot_files`, the `foot_file` add-on composed with `personal-standard` and `supplies-standard`.

Not in scope, left to siblings: nail files and clippers for fingers/toes (sibling `nail-tools`); general-purpose body sponges and loofahs, including soft foot-scrub pads that are not abrasive callus tools (sibling `loofahs`); replacement heads/pads/cases for beauty tools generally, where not specific to a foot file (sibling `beauty-tool-accessories`). Where a single listing bundles a foot file with a pumice stone (very common — kits sell both together), this register treats it as one product with two abrasive_material values rather than two registers, per the brief.

Quote labels: **verbatim** = I read the sentence myself in raw text (curl, with tags stripped, or the WebFetch tool returning the sentence inside quotation marks after I confirmed it against the raw fetch); **fetched excerpt** = the fetch tool summarised or lightly reworded the page and I did not independently confirm the exact wording; **search excerpt** = from a search-engine summary with no page read. No search-excerpt-only number is used in category.json. WebSearch reported its budget exhausted on the first three queries of this run (APMA/diabetes, cosmetology-board pumice disinfection rules, Amope roller mechanism), so everything below came from direct curl fetches, WebFetch on specific URLs, and the openFDA API. I could not reach: state cosmetology board regulation text (the specific claim that pumice stones must be single-client-use in salons because they can't be disinfected — this is widely repeated in trade and consumer press but I could not fetch a primary regulation to confirm it, so it is **not** in category.json; it survives only as color in this file); Mayo Clinic and foothealthfacts.org's own search and `my.clevelandclinic.org` all returned 403/404/JS-shell pages to curl and WebFetch.

## Seed

Seed run (recorded verbatim from the brief's output):
- Example: kitchen-knives `steel_hardness` — "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."
- Persona: "A barista trainer, who thinks in the variable most beginners get wrong."

1. **The foot-files equivalent of steel hardness is abrasion level (gentle to extra-coarse).** It is printed on packaging and compared like a spec, but like HRC it only makes sense as one side of a trade-off: coarser grit clears thick callus faster but has less margin for error against healthy skin, the same shape as "harder steel holds an edge longer but chips." Survived the data check: Amope's own FAQ names three roller grades for exactly this reason (Extra Coarse, Coarse, Soft Touch) and the existing add-on already has an `abrasion_level` enum. **Kept** as the `abrasion_level` facet, with a trade-off entry and a scale visual grammar borrowed from Amope's own roller naming.
2. **The barista trainer's variable most beginners get wrong: beginners price-compare "foot file" against "pumice stone" as if they were graded on the same axis, when the variable that actually controls the outcome is whether the skin is wet or dry when you use it and whether the device needs replaceable heads.** A barista trainee obsesses over bean origin while missing grind size, the actual lever on extraction; a first-time foot-file buyer obsesses over price and material name while missing that pumice must be wet to work and powered rollers must be dry, and that a $15 roller is really a $15-plus-refills purchase. **Kept**: wet-versus-dry use folded into the `abrasive_material` facet's definition (each value states whether it's a wet or dry tool) rather than a separate split, because the answer is fully determined by material choice and a separate question would just restate it. The refill-cost point survives as the `refill_dependency` facet and the `ongoing_refill_cost` derived metric. **Dropped**: making "just for you, or shared with others?" its own split. I considered it (porous pumice raising hygiene questions in a shared bathroom is genuine category knowledge), but the question itself reads unchanged for combs, razors or toothbrushes, and I could not source the specific salon/cosmetology regulation that would make it more than a generic "is this shared" question (see "Could not reach" above). It survives as the `porous_material` annotation and a trade-off, not a split or a judgement.

## Sources and what each supports

### Standards and professional bodies (podiatry)

**American Podiatric Medical Association (APMA), Diabetes page**
https://www.apma.org/diabetes/
- verbatim (read from the raw fetched HTML with tags stripped): "Even a simple callus can progress rapidly to a non-healing, limb-threatening wound in a patient with diabetes."
- Supports: the diabetes/neuropathy caution judgement. Confidence high. This is APMA's own patient-facing page, not a search summary; I located it by crawling `apma.org`'s home page links, not by search.

**American College of Foot and Ankle Surgeons (ACFAS), FootHealthFacts.org, "Callus"**
https://www.foothealthfacts.org/conditions/callus
- verbatim (read from raw fetched HTML, after finding the real article URL via the site's own `sitemap.xml` — the page itself is JS-rendered and the body text is embedded in the HTML behind a "Javascript is required" fallback block, which is where I read it): "People who have calluses are cautioned against performing "bathroom surgery," as this can lead to cuts and infection. A foot and ankle surgeon can evaluate the cause of the calluses and recommend the treatment most appropriate for your condition."
- Supports: the general self-treatment-risk judgement and the `bathroom surgery` vocabulary term. Confidence high.

**ACFAS, FootHealthFacts.org, "Diabetes Complications and Amputation Prevention"**
https://www.foothealthfacts.org/conditions/amputation-prevention-diabetic
- verbatim (same JS-fallback-block method): "Neuropathy causes loss of feeling in your feet, taking away your ability to feel pain and discomfort, so you may not detect an injury or irritation." … "Corns and calluses. When neuropathy is present, you cannot tell if your shoes are causing pressure and producing corns or calluses. Corns and calluses must be properly treated or they can develop into ulcers." … "No "bathroom surgery." Never trim calluses or corns yourself, and do not use over-the-counter medicated pads."
- Supports: the diabetes/neuropathy judgement, specifically and explicitly naming self-treatment of calluses (not just corns/blades) as the thing to avoid. Confidence high. Two independent ACFAS pages plus APMA (above) agree, so I treat this as sourced twice over by two bodies, not one body twice.

### Specialist / primer

**Healthline, "Corns and Calluses" (read with curl + tag-stripping, not WebFetch's summary)**
https://www.healthline.com/health/corns-and-calluses
- verbatim: "It may also help to soak corns or calluses in warm water for 20 minutes. Then gently rub the corn or callus with your finger or pumice stone."
- verbatim: "Use a foot file or pumice stone to remove patches of hard skin from your feet. Replace your foot file regularly. Allow your pumice stone to dry thoroughly between each use."
- Supports: pumice/files used after soaking (the general-advice side of the wet/dry disagreement below); "replace regularly" and "dry thoroughly between uses" as the practical hygiene behavior behind the `porous_material` facet and the pumice trade-off. Confidence high (I read the raw text myself).

### Brand / spec sheet

**Amope, "Pedi Perfect™ Electronic Foot File FAQs"**
https://www.amope.com/en-us/pedi-perfect
- verbatim (raw HTML, tags stripped — this is the brand's own FAQ page, not a marketplace listing): "The device roller bears microgranulated particles that ensure the gentle and easy removal of hard skin without damaging the intact skin layers." … roller types are "Extra Coarse, Coarse, and Soft Touch." … "No, feet should not be soaked in a footbath before the application of Amopé Pedi Perfect™ Electronic Foot File with Diamond Crystals. … An optimal removal of hard skin is carried out on dry skin." … "If you apply too much pressure, the product automatically switches to OFF. This built-in safety stop protects your skin." … "The rollers must be replaced whenever they are worn out."
- Supports: the `power_type`, `abrasive_material` (diamond_crystal_roller value), `abrasion_level` (three named grades, used for the scale visual grammar), the dry-only side of the wet/dry disagreement judgement, the pressure-cutoff safety feature (trade-off 1), and `refill_dependency` / the `ongoing_refill_cost` derived metric. Confidence high.

**PedEgg (BulbHead), product site**
https://pedegg.com/
- verbatim: "Stainless-steel microfile · Collection cup · 2 emery pads." … "The stainless-steel microfile helps remove dry skin and calluses." … "Do not use on broken skin or remove too much skin."
- Supports: the `stainless_steel_rasp` and `emery_paper` values of `abrasive_material`, the dry-use metal-rasp family, and a second, independent "do not use on injured skin" caution alongside Amope's (both makers warn against their own device on broken/injured skin, which I read as the maker's side of the same caution the podiatry bodies state more strongly). Confidence high.

### Primer / reference

**Wikipedia, "Pumice"**
https://en.wikipedia.org/wiki/Pumice
- fetched excerpt (WebFetch, quoted spans): "Pumice is a volcanic rock that consists of extremely vesicular rough-textured volcanic glass, which may or may not contain crystals." … "pumice is widely used as a skin exfoliant" … "It was discovered in a Roman poem that pumice was used to remove dead skin as far back as 100 BC."
- Supports: the mechanism description of pumice as porous (vesicular) volcanic rock, and the historical color in the mechanism section. Confidence high for what pumice is (basic geology, unlikely to be wrong even as a fetched excerpt); the 100 BC claim is attributed to "a Roman poem" per the article itself, so I keep it as color, not a hard date.

**Wikipedia, "Nail file"**
https://en.wikipedia.org/wiki/Nail_file
- fetched excerpt: "Glass nail files … have a smoother and more even surface, they do not splinter the nail like emery boards or metal nail files." … "This makes them a preferred instrument by manicurists."
- Supports: the `glass_ceramic` value's non-porous, continuous-surface description, and why manicurists/pedicurists prefer glass. Does not itself state a porosity or disinfection claim (I checked and the article doesn't make one), so the `porous_material` facet's "glass/ceramic/metal are sealed" side rests on the material being solid ground glass (no pores by definition) rather than a direct quoted hygiene claim — this is a lower-confidence inference, flagged in the judgement below.

### Brand / spec sheet, added post blind-check (2026-10-25)

WebSearch had already used its session budget when I started this fix pass (it reported this on the very first query, same as the original research pass), so everything below came from direct `curl` fetches (desktop user agent) to URLs found by crawling drscholls.co.uk's own collection pages, not from search.

**Dr. Scholl's (Scholl's Wellness Company), "Nano Glass Foot File" product page**
https://drscholls.co.uk/products/nano-glass-foot-file
- verbatim (curl, tags stripped): "WASHABLE, RUSTPROOF AND BUILT TO LAST: Waterproof, reusable and easy to clean, this foot file is hygienic and rustproof for long-term use." … "After use, rinse the foot file under running water and let it dry."
- Supports: a real, sourced "glass files rinse clean" claim for a glass *foot* file (not the Wikipedia nail-file page used before, which was about fingernails and carried its own `[citation needed]`). Confidence high. Fixes required-fix #2.

**Dr. Scholl's, "Pedi Perfect Pro Electric Foot File" product page**
https://drscholls.co.uk/products/pedi-perfect-pro
- verbatim: "Featuring a 90-minute rechargeable battery, dual-speed motor, and waterproof design" … "90-MINUTE RUN TIME & RECHARGEABLE: Long-lasting rechargeable battery delivers up to 90 minutes of continuous use on a single charge with an easy-to-read LED battery indicator" … "Adjustable motor with two speed settings... providing gentle smoothing or a more intense exfoliation" … "innovative multi-directional nanotech roller heads."
- Supports: a real source for the rechargeable value of `battery_type` (fixes required-fix #1, one of two sources); a second brand's roller-head mechanism, strengthening the `makes_irrelevant: abrasive_material` branch leg (fixes required-fix #6); a second brand naming a gentle-to-intense abrasion spectrum, used as the second source for `abrasion_level`. I did not use the specific "90-minute" figure anywhere in category.json, since it has only one source and the house rule needs two for a number. Confidence high for what's quoted.

**Dr. Scholl's, "Electric Foot Callus Remover" product page (Ireland storefront)**
https://drscholls.co.uk/products/ie-callus-remover-electric-foot-file
- verbatim: "Cordless and lightweight... 4 x AAA batteries included." … "Replace all batteries at once; do not mix old and new or us[e] rechargeable batteries." … "Use on dry skin only... Make sure the roller head is securely fitted."
- Supports: a real source for the disposable_battery value of `battery_type` (the other of the two sources for fix #1), and a second, non-"Pedi Perfect"-branded confirmation of the roller-head mechanism (fix #6). Confidence high.

### Regulator (checked directly, not via search)

**openFDA, device classification database**
https://api.fda.gov/device/classification.json
- Queried directly for `device_name:"callus"` and `device_name:"foot file"`; both returned `{"error":{"code":"NOT_FOUND"}}`.
- Supports: the judgement that no FDA device classification exists for these products, consistent with this being an unregulated cosmetic/wellness category rather than a medical-device one. Confidence medium — absence of a hit is weaker evidence than a hit, and I did not check every plausible search term (e.g. "callus remover," "corn remover").

## Data points used for derived metrics

No sourced per-unit prices or refill-pack counts were found (brand pages don't list refill pack sizes or prices on the FAQ/spec pages I could fetch; marketplace listings that would carry them blocked the fetch). The `ongoing_refill_cost` derived metric is therefore given only as a formula over facet/listing fields (`refill_pack_price / rollers_per_pack`), with no invented numeric example, per the "do not invent numbers" rule.

## Against the existing add-on schema

Add-on read: `supplies/personal/add-ons/foot_file.jsonc` — currently a single field, `abrasion_level` (enum: gentle, medium, coarse, extra_coarse, unspecified), composed with `personal-standard` (`replacement_compatibility_raw`, `sizing.weight_capacity`) and `supplies-standard` (`high_signal_markdown`, `sizing`).

| Facet in this doc | In add-on today | Proposed kind under AGENTS.md | Note |
|---|---|---|---|
| abrasion_level | yes, exactly this enum | category quality currency | Keep as is; this register's own grades (gentle/medium/coarse/extra_coarse) match the existing enum, and Amope's three-grade roller naming maps onto it with "Soft Touch" = gentle/medium and "Extra Coarse" = extra_coarse. No change proposed. |
| power_type | no | hard filter | Buyers search "electric callus remover" vs "foot file" as different product types. Usually in the title; cheap to extract. Proposal: a `power_type` enum [manual, battery_powered] on the add-on. |
| abrasive_material | no | hard filter | Buyers also search by material ("glass foot file", "pumice stone"). Usually in the title, sometimes a bullet; rarely a dedicated spec field, so `source: listing` with low backfill cost. Proposal: an `abrasive_material` enum [pumice, glass_ceramic, stainless_steel_rasp, emery_paper, diamond_crystal_roller]. |
| battery_type | no | likely noise as a standalone field | Only applies to the `battery_powered` branch, so it is null for most of the catalog — AGENTS.md's density test for a top-level field fails. Recommend folding into `high_signal_markdown` rather than adding a column, unless the powered sub-catalog is large enough on its own to justify one. |
| porous_material | no, and shouldn't be one | not a stored field | Fully derivable from `abrasive_material` (true for pumice/emery_paper, false otherwise); storing it would just duplicate the material field. Keep it computed, not stored. |
| refill_dependency | partly — `replacement_compatibility_raw` (from `personal-standard`) already exists and is the right field | existing field, no change | This register's facet is this field under a different name; no new schema field needed. The pipeline should populate `replacement_compatibility_raw` with the roller-head SKU/name for powered units and leave it null for manual tools. |

No add-on field is proposed for removal; the add-on currently has only the one field and it is kept.

## Branches

- `power_type` = `battery_powered` → `then_ask: [abrasion_level, battery_type]`, `makes_irrelevant: [abrasive_material]`. Traced to the Amope FAQ: powered rollers are sold by grit grade (Extra Coarse/Coarse/Soft Touch) and by battery type; the question "pumice, glass, or metal rasp?" has no answer once the buyer has already said "powered roller," since every powered unit in the sources I read uses a mineral/diamond-coated roller, not a hand-filed material.
  - **Updated 2026-10-25, post blind-check**: `then_ask: battery_type` now traces to two real sources instead of none — Dr. Scholl's Pedi Perfect Pro (rechargeable, LED charge indicator) and Dr. Scholl's basic Electric Foot Callus Remover (disposable AAA batteries, explicitly "do not... use rechargeable batteries"). See `drscholls_pedi_perfect_pro` and `drscholls_callus_remover_aaa` below. `makes_irrelevant: abrasive_material` is also stronger now: both Dr. Scholl's powered products use a "roller head," confirming the roller-not-hand-filed-material pattern holds across a second brand lineage, not just Amope. Flagged honestly: Scholl's Wellness Company (UK) and Edgewell's Amopé (US) share "Pedi Perfect" branding history in some markets, so this is not a fully unrelated third-party confirmation — I treat the AAA-battery basic model, which does not carry the "Pedi Perfect" name, as the more independent of the two.
- `power_type` = `manual` → `then_ask: [abrasive_material]`. Traced directly to the split rationale: the manual branch is exactly where material choice (pumice vs glass vs metal rasp vs emery) is still an open, meaningful question. Unchanged by the blind check.
- No branch is attached to `abrasive_material` itself. I checked whether choosing pumice vs. glass vs. metal rasp changes what's asked next (abrasion_level) and it does not — all four manual materials come in a gentle-to-extra-coarse range per the sources above, so the next question is the same regardless of which material was picked. A branch here would not trace to anything in the register, so I dropped it rather than force one, per the "only write a branch where the answer really changes the path" rule.

## Trimmed detail

- The Amope FAQ item on correct hand position ("Clasp the Electronic Foot File with one hand at the grip and move it gently over the hard skin") and the "how often can I use it" item (no fixed frequency given, "as often as you like") were read but don't support any facet or judgement in the JSON; they're usage instructions, not buying criteria.
- The PedEgg page's cleaning instructions ("wipe the tool with a damp cloth and allowing it to air dry before storage") were read but not used as a sourced number or threshold; they support the general "dry thoroughly" hygiene point already carried by the Healthline quote, so I didn't double-cite.
- I looked for, but could not source, a specific salon/cosmetology-board rule that pumice stones must be single-client-use because they can't be disinfected. This is widely repeated in consumer press and is plausible given pumice is porous (sourced) and porous tools are generally harder to disinfect (general knowledge, not from a regulation I read), but I would not put a specific regulatory claim in category.json without reading the regulation itself, so it stays out.

## Corrections

None yet — this is the first pass, run end to end before any blind check. If a blind-check pass (per `project-category-expertise-outputs` conventions) finds issues, log them here rather than silently editing the JSON.

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

All 8 required fixes from `check.md` were addressed; none rejected.

1. **`battery_type` was wholly uncited (check.md #1).** Fetched two new sources — Dr. Scholl's Pedi Perfect Pro (rechargeable, LED indicator) and Dr. Scholl's basic Electric Foot Callus Remover (disposable AAA batteries, explicitly incompatible with rechargeables). Replaced `amope_pedi_perfect_faq` with `drscholls_pedi_perfect_pro` and `drscholls_callus_remover_aaa` in the facet's `evidence`, reworded `why_experts_care` to drop the unsourced "the internal cell eventually degrades" detail and the unverified "USB" specifically, and updated `listing_field_hint`. Did not invent or carry over any number (e.g. "90 minutes") since it has only one source.
2. **Mechanism's "glass and ceramic files... don't splinter and rinse clean" had no real support (check.md #2).** Fetched Dr. Scholl's Nano Glass Foot File page, a real glass *foot* file (not the fingernail-file Wikipedia page used before). It directly supports "rinse clean" and "rustproof," not "don't splinter," so I replaced "don't splinter and rinse clean" with "rinse clean and don't rust" in `mechanism.plain` and the matching family entry, and added `drscholls_nanoglass` to `mechanism.evidence`.
3. **`abrasive_material`'s cold-start question dropped `emery_paper` (check.md #3).** Changed the question (and matching `cold_start_question`) from "Natural pumice, glass or ceramic, or a metal rasp?" to "Pumice, glass or ceramic, metal rasp, or emery paper?" — now one option per manual-branch value, still 9 words.
4. **`abrasion_level`'s `backfill_cost: low` contradicted its own `listing_field_hint` (check.md #4).** Raised to `medium` and rewrote `backfill_note` to say the field is low-cost only on the powered branch (printed on roller-head packaging) and needs visual inspection on manual pumice/glass listings. Added `drscholls_pedi_perfect_pro` as a second evidence source (a second brand naming a gentle-to-intense spectrum).
5. **`refill_dependency` was marked `researched` but is mostly `derived` (check.md #5).** Reclassified `source` to `derived` and reworded `backfill_note` to say the boolean itself is derived from `power_type` for almost every row, and only the exact refill SKU/name — already tracked separately in `replacement_compatibility_raw` — needs genuine research.
6. **`battery_powered → makes_irrelevant: [abrasive_material]` generalized from one brand (check.md #6).** Fetched two Dr. Scholl's powered products; both use a "roller head," confirming the pattern across a second brand lineage. Logged in the Branches section above, with an honest flag that Scholl's Wellness Co. and Amopé share "Pedi Perfect" branding history in some markets, so I lean on the non-"Pedi Perfect"-branded basic model as the more independent confirmation. No JSON schema field carries branch-level evidence, so this strengthening lives in notes.md and in `battery_type`/mechanism evidence, not as a new JSON field.
7. **Two single-source facets didn't fully support their claims (check.md #7).** (a) `abrasion_level`'s four-value scale doesn't map 1:1 onto any one brand's named grades — added a judgement entry saying so at confidence medium, rather than asserting the Soft Touch→gentle/medium split as fact. (b) `porous_material`'s "false" half (glass/ceramic/steel/roller = sealed) rested on `healthline_corns` alone, which only covers pumice. Broadened its `evidence` to `wikipedia_pumice`, `wikipedia_nail_file`, `pedegg_site`, `drscholls_nanoglass` and `amope_pedi_perfect_faq`, and reworded `backfill_note` to say the derivation rests on each material's own sourced, solid-vs-porous description rather than a single standalone porosity study.
8. **Uncited "$5 stone vs $20 device" price example (check.md #8).** Removed the dollar figures from `power_type.why_experts_care`; replaced with "Refills can make a cheap-looking roller cost more than a stone over time" — the point survives without an invented number.

## Script notes

`validate.py`, `render.py`, and `viewer.py` ran clean on the first attempt (see command output); no script issues to report.