Teste rápido
Guess first, no penalty. You paste a user story into an AI tool and ask for test cases. Which extra request gives you the most useful result?
A user story is a short description of a feature from the user's point of view. Acceptance criteria are the conditions it must meet to count as done. AI turns both into a first set of test cases in seconds, a good first job to hand over.
Anthropic's study Labor market impacts of AI (March 2026) found Claude is used for 51.9% of the core tasks of software quality assurance analysts and testers, in its Figure 3. "Covers a task" means AI is used for that task, not that the job disappears.
What makes the prompt (the message you send to the AI) work:
- The story and its criteria, pasted in full.
- Context: web or mobile, the users, what changed.
- The cases you want: happy path, negative cases, boundary values (the values at the edge of what's allowed) and mistakes a real user might make.
- A format: a table with steps, data, expected result, priority.
- The gap check: ask it to list unclear or missing criteria.
The last line is the quick win: it often turns up a question worth asking.
A worked example
Write test cases for this user story and its acceptance criteria.
Context: web app, logged-in customers, password reset is new this release.
Include: happy path, negative cases, boundary values, and 3 cases a
real user might hit by accident.
Format: a table with ID, steps, test data, expected result, priority.
Then list any acceptance criteria that are unclear or missing.
[paste the story and acceptance criteria]Preencha as lacunas
Fill the gaps.
Prompt
Write test cases for this story. Include . Format: . Then list .
Try it
Seus passos, cerca de 2 minutos
- 1.Run the prompt on one story you're testing this week. Take one unclear criterion to the product owner. Save the prompt as the first piece of your "AI test kit".2 min