# Auteur direction brief

Use this before substantial AI-assisted product work. If an answer is unknown, write the question instead of allowing the system to invent one.

## 1. Outcome

What observable change should exist when the work is done?

## 2. User and problem

Who is affected, in what situation, and what evidence supports the problem?

## 3. Current state

How does the product behave now, and where is the relevant source of truth?

## 4. Scope

What is included, excluded, and deliberately deferred?

## 5. Constraints

What behavior, data, policy, interface, accessibility requirement, cost, or dependency must be protected?

## 6. Unknowns and risks

What needs investigation? What could make the chosen direction wrong?

## 7. Acceptance evidence

Which tests, observations, product metrics, and independent review angles will justify completion?

## Handoff

- Changed:
- Why:
- Verified:
- Still uncertain:
- Next bounded action: