Know what every dish costs before you plate it: 10 AI skills for recipe costing
waste-adjusted-yield-cost
cost a dish on what you actually get after trim and cooking loss, not the raw weight
How the two work together
Claude thinks it through. Paste the Claude prompt into Claude Code, or drop the folder into your skills folder. Claude does the judgement: what to look for, what is worth doing, what is right.
Codex gets it done. At the hand-off point Claude runs Codex on your machine with one command and passes it the Codex prompt. Codex does the mechanical part and hands the result back. Claude checks it before you see it.
No API key to set up: Claude calls the Codex you already have installed. If Codex is not installed, Claude does that half itself and tells you.
Prompt for Claude
--- name: waste-adjusted-yield-cost description: Costs a dish on what you end up with after trim, peel and cooking loss, not the raw weight bought. It turns a purchased quantity and measured usable yield into a true cost per usable kilogram, keeping the waste percentage visible so a jump in trim loss shows before it eats the margin. Use whenever a dish uses a bulk ingredient losing weight before the plate, and to check why a real cost drifted from the recipe card. --- # What a kilogram costs once the bit you can't serve is gone Give this a purchased quantity and price, and the usable weight recovered after trimming, peeling, boning or cooking it. Get back the true cost per usable kilogram, the waste percentage, and a flag if it has moved against the recipe card. ## What it does 1. **Record the purchased weight and its VAT-exclusive price**, from an invoice, as dish-cost-builder does; a guessed price inherits into every dish using it. 2. **Record the usable weight recovered**, from a real trim test, not a textbook yield. WRAP put UK sector food waste at "920,000 tonnes... wasted at outlets each year", costed at "£2.5 billion per year in 2011", and a more recent figure puts edible waste at "£0.97 per restaurant cover". Trim loss on one ingredient is a small, local version of that number. 3. **Waste percentage**: (purchased weight minus usable weight) divided by purchased weight. 4. **True cost per usable kilogram**: price divided by usable weight, not purchased weight. This is the figure that belongs in a recipe cost. 5. **Compare against the recipe card's assumed yield**, if one exists, and flag any gap. A fish yielding 48% instead of a usual 55% means a different supplier, cut, or generous cutting, worth knowing before it shows as a smaller margin with no obvious cause. ## Then it checks 1. Purchased and usable weights are both real measured figures, not assumed percentages. 2. Waste percentage and cost per usable kilogram are shown as calculations. 3. Any recipe-card assumed yield is named "assumed", never confused with the measured figure. 4. A gap of over 5 points from a prior measurement or the recipe card is flagged by name. 5. The purchase price used is confirmed VAT-exclusive, consistent with dish-cost-builder. Any check fails: redo once. Failed twice: say what is wrong and stop. ## Rules - Public information only. Never invent a fact, number or quote, or a yield percentage; it comes from a weighed test, or isn't used. - Anything sent in someone's name says whose name it is. - Never state a UK-wide average as applying to one kitchen; sector figures are context, not a target. - Never quietly replace a measured weight with an assumed one; flag the gap. - Working document for the owner to check against their trim tests and recipe cards, and their accountant on cost of sales. Not accountancy advice. ## Built from - WRAP, "Overview: Waste in the UK Hospitality and Food Service Sector", https://www.wrap.ngo/resources/report/overview-waste-hospitality-and-food-service-sector, published 1 November 2013, read 4 November 2026: the 920,000 tonnes/£2.5bn sector figures, step 2. - Public Sector Catering, "WRAP calls on people to feed the bottom line, not the bin", https://www.publicsectorcatering.co.uk/in-depth/wrap-calls-people-feed-bottom-line-not-bin, published 9 February 2023, read 4 November 2026: WRAP's £0.97-per-cover figure, Eleanor Morris, step 2.
Prompt for Codex
# waste-adjusted-yield-cost ## You are given For each ingredient: purchased quantity and VAT-exclusive price (ideally from the same invoice as dish-cost-builder), the measured usable weight from an actual trim, peel, bone-out or cook test, and the test date. Where it exists, the recipe card's assumed yield percentage. Any previous yield test on file for the same ingredient and supplier. ## Produce Write into a `./waste-adjusted-yield-cost-output/` folder: 1. `yield-tests.csv`: `ingredient_name`, `test_date`, `purchased_weight`, `purchased_unit`, `price_ex_vat`, `usable_weight`, `usable_unit`, `waste_percent`, `cost_per_usable_unit`, plus matching `_formula` columns. `waste_percent` = `(purchased_weight - usable_weight) / purchased_weight * 100`. `cost_per_usable_unit` = `price_ex_vat / usable_weight`. 2. `yield-drift.csv`: `ingredient_name`, `current_waste_percent`, `comparison_waste_percent`, `comparison_source` (`recipe card assumed yield`, `previous test on file`, `none available`), `drift_percentage_points`, `flagged` (`yes` when drift exceeds 5 points either way). 3. `gaps.md` - numbered list of: every ingredient with no comparison available; every `flagged: yes` row; any price not confirmed VAT-exclusive; any usable weight exceeding purchased weight, always an input error, never passed through silently. ## Rules - Codex calculates, compares and flags. It never invents a yield, trim loss or usable weight not measured and supplied. - `waste_percent` and `cost_per_usable_unit` are always shown with their formula. - A missing comparison is recorded as `none available`, never filled in with an assumed industry figure. - An input where usable exceeds purchased weight is listed in `gaps.md` as an input error and excluded from `yield-drift.csv` until corrected. - Never write a UK-wide average into any ingredient's own row. - British English, GBP, DD Month YYYY dates. No em dashes. - Every file ends with: this is a working document prepared for the owner to check against their own weighed trim tests and recipe cards, and for their accountant where the figures feed into cost of sales. It is not accountancy advice. ## Return The absolute path and row count of each file. For every ingredient: purchased weight, usable weight, waste percentage, and cost per usable unit, with formulas. Every flagged ingredient, with drift and comparison source. The number with no comparison available. Any input error found. The `gaps.md` item count.
Built from the best public work on this
Sources for waste-adjusted-yield-cost
Everything below was opened and read on 4 November 2026. Nothing is cited that could not be loaded.
1. Waste and Resources Action Programme (WRAP), "Overview: Waste in the UK Hospitality and Food Service Sector"
https://www.wrap.ngo/resources/report/overview-waste-hospitality-and-food-service-sector, published 1 November 2013, read 4 November 2026.
This is WRAP's own archived report page, and it states the sector figures used in step 2 directly: "920,000 tonnes of food is wasted at outlets each year", with the cost "estimated at £2.5 billion per year in 2011." The page itself notes that WRAP has since published more recent figures, dated January 2020, which this pack has not separately verified.
These numbers are used only to show that trim and cooking loss on a single ingredient is a small, local instance of a cost that is large and real across the sector; they are explicitly not used as a per-outlet benchmark, since the report gives a sector total, not a per-kitchen figure.
Where the skill departs: the report covers the whole waste stream, including plate waste and spoilage, not specifically ingredient trim loss. The skill isolates trim/yield loss as the one piece a recipe cost can and should capture directly.
2. Public Sector Catering, "WRAP calls on people to feed the bottom line, not the bin"
https://www.publicsectorcatering.co.uk/in-depth/wrap-calls-people-feed-bottom-line-not-bin, published 9 February 2023, read 4 November 2026.
Quotes WRAP's Eleanor Morris directly: "This edible food waste is estimated to cost £0.97 per restaurant cover, on average." This is the most recent, specific, attributable figure found for this pack, and it is a per-cover figure rather than a sector total, making it a closer analogue to what a single kitchen might measure.
Where the skill departs: this is a trade-press article reporting a WRAP spokesperson's figure, not WRAP's own underlying report, which was not located during this research. The figure is used as context for why yield and waste matter financially, never presented as this skill's own measured output or as a target any individual kitchen should expect to hit.
Best public prompt we found for this job
The closest public artefact is `SpandanPriyam22/FeedForward` on GitHub, read from api.github.com on 4 November 2026, which returned a `stargazers_count` of 19 and a description opening "Food Waste Management web application" (the full text, paraphrased for length, covers monthly statistics on food wastage, costs, quantity purchased and CO2 emissions, a personal inventory, and a map-based donation feature). It is the best-reviewed repository found for anything in this food-waste space.
What we took: the principle of tying a waste figure to both a quantity and a cost in the same record, rather than logging waste as weight alone with no financial consequence attached.
What we did not copy, and why: it is a consumer food-waste and donation-matching app, tracking what gets thrown away after purchase, with no concept of yield at all, that is, the planned, structural loss between a raw ingredient and the usable product a recipe is actually built from (trim, bone, peel, evaporation on cooking). A restaurant's biggest yield cost is usually not waste in the bin sense; it is the weight that was never going to be served in the first place, and that calculation, and the drift-detection against a recipe card's assumed yield, is not something this kind of tool is built to do.
Want this running in your business?
I optimise how businesses run — your sales, your visibility, your social media — and build bespoke software where nothing off the shelf fits. The first conversation is free. Work starts from £150 a day.
Foxera