Work orders against bills of materials for a small manufacturer, with availability checks, stock moves, WIP, and a dashboard
work-orders-bom · 3600 s per turn · source
The ask
We make stainless steel work tables, carts, and shelving on three production lines, and I need to run our work orders against our bills of materials. Before a work order starts, the app has to check that we have enough of every component on the shelf; starting it takes those parts out of stock, and finishing it puts the finished units into stock. Our parts list, bills of materials, and work orders are attached; they came out of spreadsheets and aren't clean, so check them rather than trusting every row. I want team logins with different access: each line supervisor sees and runs only their own line's work orders, and our accountant can see everything but change nothing. Give me a dashboard showing what's open and what we have tied up in work in progress, an admin login for me, and a login for Sam Whitaker, who supervises Line 2.
Seed files the agent must import: parts.csv</code>, <code>boms.csv</code>, <code>work_orders.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, 21, 24.
| Tag | Items |
|---|---|
| Exact | 9 |
| Permission | 3 |
| Rule | 3 |
| Automation | 2 |
| Dashboard | 2 |
| Delivery | 2 |
| Sharing | 2 |
| API | 1 |
| Audit | 1 |
| Persistence | 1 |
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
Give me a monthly production report by line: how many units each line finished and what those units cost us in materials. And put our on-time completion rate on the dashboard.
Change 2 · 3 new checklist items
When a part drops below its reorder point, raise a purchase request for it automatically, for enough to bring it back up to twice the reorder point, and never a second one while the first is still open. When we receive the parts, put them into stock and close the request.
Change 3 · 3 new checklist items
Add a Quality check step between In progress and Completed. Only our QA lead can pass a work order there, finished units go into stock only for the units that pass, and the rest get recorded as scrap.
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.