How to use Cursor to draft user stories from code
A subject expert used Cursor to turn detailed project context, existing code, and infrastructure information into about 45 user stories, then refined the set through individual and team review.
news.ycombinator.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Using [project context], [related projects], and [tagged implementation files], draft user stories for [intended work]. Retrieve relevant information using [authorized infrastructure CLI]. Follow our existing styles, identify missing context, and prepare the stories for my review.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source poster / publisher
- wincy
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- The conversation generated about 45 stories. After individual and coworker review, the team removed two or three and corrected one or two; the author reported spending half a day.
- Human involvement
- The author reviewed the output for hours, then worked with coworkers to remove and correct stories.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Add related projects and tag files showing similar implementations.
- Explain system quirks, project goals, and information locations.
- Ask Cursor to retrieve relevant infrastructure information through the CLI.
Asked to write user stories for a complex work project, the author spent two hours giving Cursor context that it would otherwise lack. They added related projects to the workspace, tagged files containing similar implementations, described system quirks, and explained where relevant information could be found. The preparation drew on their subject expertise and experience doing the process on two earlier projects.
They then asked Cursor to use the CLI to retrieve relevant infrastructure information and write stories describing the intended work. The conversation generated about 45 stories. The author spent another few hours reviewing them before taking them into a meeting with coworkers.
That discussion led to deleting two or three unnecessary stories and correcting one or two affected by missing context about page coloring. The author reported spending half a day overall and found the tool particularly helpful for turning knowledge into organized scrum stories. Their estimate that the work might otherwise take a week or two was a personal counterfactual, not a measured comparison. Who it's for: subject experts who can supply detailed context and lead a substantive review of generated planning documents.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · wincy
Frequently asked questions
What context did the author supply before requesting stories?
They added related projects, tagged relevant implementation files, described system quirks, and explained where to find information. They spent two hours on this preparation.
How were the generated user stories reviewed?
The author reviewed them for several hours and then discussed them with coworkers. They deleted two or three stories and corrected one or two with insufficient context.
More details
Remaining steps
- Generate user stories describing the intended implementation work.
- Review the stories individually and discuss them with coworkers.
- Delete unnecessary stories and correct missing contextual details.
About this account
Cursor gathered additional information and organized the supplied knowledge into stories after extensive human preparation.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=47577570. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: wincy on Hacker News
Source published
Added to Agents In Use
Source checked
Tools and topics
Cursor · CLI · Opus 4.6
User stories · Project planning · Requirements
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.