Write user stories with AI
Turn a feature request into a user goal and a reason it matters, then give the team acceptance criteria it can test against the working product.
Write user stories with AI when a backlog item says “add an export” but nobody can explain who uses it or what they will do with the file. The writing-user-stories skill asks an agent to state the actor, the action and the benefit. That might turn a vague export into a finance analyst’s need to reconcile approved transactions for a selected period.
Acceptance criteria then give the team something observable to test. An empty date range or a user without access may deserve its own case. A story can be clear about the result while leaving room for the team to choose how the implementation works.
What it does
- Identify the user whose work the feature changes.
- State the benefit in terms that a product owner can discuss.
- Write criteria covering a starting condition, action and observable result.
- Split an oversized story into useful, testable slices.
- Review the result against the INVEST qualities, including independence, value and testability.
Install and use
- Provide the feature request, the intended user and any known constraints. Drafting needs no project-tracking account.
- Ask, “Turn the monthly finance export into small user stories and include an acceptance case for a period with no approved transactions.”
- Review the stories with the team and rewrite criteria that depend on an undefined judgement such as “works well”.
FAQ
Source
Made by Britt Crawford. Original skill: writing-user-stories, published on GitHub under the MIT licence.
View on GitHub