FOR PRODUCT LEADS AND DECISION-MAKERS
“It’s done.” Does it do what we wanted?
The ticket is closed, the pipeline is green, and in the demo it turns out partial refunds don’t work at all. FACET keeps what you agreed on and shows what of it is built and tested — before you sign off.
checkout/refunds/partial-refund
Partial refund
| A refund is promised within 14 daysrefund-within-14-days | agreed |
|---|---|
| A full refund returns the whole amountfull-refund-returns-total | named ×1 |
| A partial refund returns only the chosen itempartial-refund-returns-item | open |
ONE SHARED YARDSTICK
Product means one thing. Engineering builds another.
Product asks: does the feature keep what we promised? Engineering asks: what exactly should I build and test? With FACET both read the same sentence — given, when, then — instead of two documents that contradict each other.
Product
“Does it keep our promise?”
contract: scenarios: - id: partial-refund-returns-item given: an order with two items was paid when: the customer returns one item then: only that item’s price is refunded
Engineering
“What exactly must we evidence?”
Illustrative example · one scenario from a feature record
THE QUESTION FOR ENGINEERING
One question is enough to see where you stand.
No documentation project. Ask your engineers about one important feature: where is it written what it should do, and which test checks it? If the answer comes quickly, good. If not, your team adds FACET to its coding agent — and next time the answer sits next to the code.
For one important feature, can we show what was agreed, tested, and left open?Take the partial refund. Which scenarios exist, which test backs which one, and what is still missing?
Ready to copy · a question, not a requirement
WHAT FACET DOES NOT DO HERE
The answer lives in the repository. The decision stays with you.
FACET does not set product priorities or decide which business rule is correct.