Breakfast shop owner-operator - the break-even and "what a hire costs" calculator: how much to sell daily to break even, one helper converted into cups of soy milk

Build a single-file HTML break-even calculator for a breakfast shop
owner-operator - one that answers "how much must I sell each month to not
lose money" and "what does each decision translate to in cups of soy
milk."

Inputs (hand-entered; all from the owner's own books, the tool preloads
no numbers):
- Fixed costs (monthly): rent, utilities and gas, total current wages,
  platform fees
- Trading days: days actually open per month
- Item table (one row each): item / price / unit cost (flour, filling,
  eggs, beans, cups and bags allocated by actual usage). An "est" flag
  is allowed - any result using an est-flagged input carries a marker
- One "anchor item" is selected - the item that best represents the
  store's revenue (e.g. buns or soy milk)

Calculations and output:
- Daily break-even line: monthly fixed costs divided by trading days
  divided by anchor-item unit margin = "how many anchor items per day
  to break even." Also give a combined view: converted by the current
  sales mix into one screen-sized statement like "N trays of buns plus
  M cups of soy milk"
- Three converters (the heart of the tool, one big line each):
  a. One more hire: proposed monthly wage divided by 30 divided by
     anchor-item unit margin = how many more anchor items per day pay
     for this person
  b. Price-raise simulator: after a [delta] raise on an item, how many
     fewer of that item per day still hold the same profit
  c. Waste translator: the cost of a binned batch divided by anchor-item
     unit margin = how many cups that tray was sold for nothing
- Sensitivity ranking: rent / utilities / wages / unit cost each +10%,
  break-even line recomputed, ranked by impact - telling the owner
  which of the four knobs hurts most

Implementation requirements:
- Single file, no external dependencies, no network calls; data in
  localStorage, survives refresh
- Every calculation shown on the page (formula plus the substituted
  numbers) - no black-box results
- Print-friendly, fits one A4 page
- No login, no cloud sync

Notes for the coding agent:
- The "est" flag must carry through to the final output: if any input
  feeding a calculation is flagged, the break-even line gets a
  "contains estimates" marker, clickable to list which inputs
- The unit-cost field's placeholder states its basis: "allocated by
  actual usage - the flour and filling cost of one tray of buns divided
  by the count, not the purchase price per unit"
- The sensitivity ranking must express impact only as numbers, no
  trend charts

Why it is written this way

Owner-operators don't fail at arithmetic; they fail at decision-scale arithmetic. "Three thousand rent, two thousand a day in sales, seems fine" is how the thinking actually runs - and it breaks down exactly at decisions like hiring: a monthly wage of a few thousand feels abstract until it converts to "eighty more cups of soy milk every single morning," at which point the counter at 6:30 a.m. answers the question by itself. All three converters fold into anchor-item counts deliberately - every decision lands in the same concrete unit, the owner can sanity-check it against the steamer trays, and if the tool is wrong, they can see where.

"Every calculation visible on the page" is what separates a tool for an owner-operator from a tool for a company. Black boxes earn no trust here - the owner cross-checks against their own rough mental number, and one mismatch kills the tool entirely. Laying out formulas and substituted numbers means that even when it's wrong, they can point at the wrong line. The "est" marker works the same way as an honest exit: nobody can fill in a precise unit cost, and demanding precision gets you an invented number that quietly poisons every later result. Allow the estimate but force the label, so at least the reader knows which parts are soft.

Putting the sensitivity ranking last is deliberate restraint. The classic misuse of a break-even tool is long-range planning - "at this line I'll save up a down payment for store two by next year" - all of which is fiction. Ranking "which of the four knobs hurts most at +10%" pins the tool to comparing the current structure, extrapolating nothing further.

Variables