Business Bench

Tasks · build · retail-back-office

Back office for two cafes with menu items, modifiers, nightly cash-ups, item performance, price-change audit, and a dashboard

point-of-sale-backoffice · 3600 s per turn · source

The ask

We own two cafes, Harbor Street and Mill Road, and I need a back office for them: our menu items with their modifiers, and each store's daily takings. Every night the manager enters the day's gross sales, discounts, card total, and the cash they counted; the app should work out whether the till is over or short and make them explain anything more than $5 off. I want to see how each item is performing, and only I change prices, with every price change on record: who changed it, from what, to what, and when. Our menu, modifiers, daily takings, and item sales since July are attached, exported from two different tills, so expect some mess. I want team logins with different access, where each store manager sees only their own store and our bookkeeper sees everything but changes nothing, plus a dashboard, an admin login for me, and a login for Owen Brooks, who manages Mill Road.

Seed files the agent must import: items.csv</code>, <code>modifiers.csv</code>, <code>daily_takings.csv</code>, <code>item_sales.csv. The agent must leave a RESULT.json with a reachable URL, an admin login, and the named restricted login.

Acceptance checklist

26 binary items. The first 14 are the enterprise baseline shared by every build task (delivery, invitations, roles, server-side authorization, scoped live dashboard, exact lists and exports, audit metadata, validation, persistence); the rest are application-specific. Core items, which together define "first usable": 1, 3, 5, 6, 8, 14, 15, 19, 23.

TagItems
Exact8
Permission5
Rule3
Audit2
Dashboard2
Delivery2
Sharing2
API1
Persistence1

The full checklist with per-item "how to check" instructions and expected figures is in checklist.md. Testers see it; the agent never does.

Change requests

Each change is a new one-shot turn on the carried-forward workspace. Only the owner's text below is sent; the tester items are withheld. After each change the full checklist is rerun, so regressions count.

Change 1 · 3 new checklist items

I want to know our average ticket for each store each month, and a day-of-week report so I can see which days sell best at each store.

Change 2 · 3 new checklist items

Any day where the cash is short by more than $20 should land on a follow-up list automatically, and stay there until I mark it resolved with a note saying what happened.

Change 3 · 3 new checklist items

Daily entries should go through Draft, then Submitted by the store manager, then Approved by me, and once I've approved one only I can change it. Also add a shift lead login that can start a draft for their store but can't submit it.

No published score. Build runs need a human acceptance pass on a live URL, and no arm has completed one for the full set. Automated probes generate a tester sheet; they do not substitute for it. Results will be published as a separately labeled release when an arm completes.