Rating
- size:
3 - complexity:
2
Purpose
Practice using AI to clarify ambiguous requirements without letting the model invent policy decisions.
Scenario
Current story:
Users should be able to pause notifications for a while if they need a break.
Stakeholder note:
This should work across email and push. We do not want people to miss critical messages. It should feel flexible.
Current acceptance criteria:
- user can pause notifications
- user can resume notifications
- system should respect user preference
Task
Use AI to:
- rewrite the acceptance criteria
- surface missing questions
- separate what is known from what still requires stakeholder decision
Expected output
- a cleaner acceptance-criteria draft
- a list of open questions
- a short note identifying which statements came from the source and which were AI suggestions
Good AI uses
- finding hidden ambiguities
- drafting clearer requirement language
- proposing question categories such as timing, exceptions, and channels
Verification focus
- no AI-generated rule should be accepted as policy without source backing
- critical-message exceptions must be explicit, not implied
- time-zone, duration, and channel behavior should be treated as open questions unless specified
Anti-patterns to watch
- turning vague input into false precision
- accepting polished wording as real agreement
- hiding unresolved business decisions inside nice prose
Debrief prompts
- what did AI clarify well
- where did AI overreach beyond the source material
- what should go back to a product owner or stakeholder before implementation begins