## Unit of Work - **Unit:** [name] - **Ticket:** [KEY-123] - **Requirements or stories served:** [IDs] ## What changed and why [Plain English: what this unit now does, and why.] ## Risk tier [Critical / Important / Standard]. Review needed: [line by line / findings-led / sampled]. ## Tests that prove it - [Test]: proves [requirement or acceptance criterion] ## Independent AI review - **Reviewed by:** [separate session or model] - **Findings:** [n critical, n high, n medium, n low] - **Resolution:** [how each critical and high finding was resolved] ## Rules changed or added [Every change to what data is shown, who can do what, a calculation, a status change or an automatic action, each with its decision reference. Write "None" if there are none.] ## Tweaks [n] cosmetic Tweaks accepted in the room (layout, styling, wording), listed in the feedback record: [link]. None of them changes behaviour. ## Feature inventory [Anything in this change that traces to no requirement, for the experts to keep, change or remove. Write "None" if there is nothing.] ## Assumptions [Each one labelled ASSUMPTION until a person confirms it. Write "None" if there are none.] ## Reviewer checklist - [ ] The change matches the requirements it says it serves - [ ] Reviewed to the depth its risk tier needs - [ ] Every rule changed or added is backed by a decision - [ ] No Tweak changes behaviour - [ ] No open QUESTION: threads