raphaelmans/agent-skills/user-story
Claude Code and agent skill by raphaelmans/agent-skills
- community
Checking that the SKILL.md file is still in the repository.
Generate structured user stories with acceptance criteria from any input (PRD, feature spec, verbal description, or codebase exploration). Use when asked to "write user stories", "create user stories", "break this into stories", "what are the user stories for...", or any request to define what users can do for a feature. Also trigger when the user mentions "acceptance criteria", "Given/When/Then", or wants to define feature behavior from the user's perspective. Do NOT use for single bug fixes, tiny config changes, refactors with no behavior change, or exploratory spikes.
Install
npx vskill@latest install raphaelmans/agent-skills/user-storySKILL.md
User Story Generation
Output Structure
Naming Rules
| Part | Format | Example | |
|---|---|---|---|
| Domain folder | lowercase, kebab-case | orders, user-management | |
| Sub-feature folder | lowercase, kebab-case | place-order, cancel-order | |
| Story file | US-{DOMAIN}-{NNN}.md (UPPERCASE domain) | US-ORDERS-001.md | |
| Domain index | index.md | index.md |
Numbering Rules
- Format:
US-{DOMAIN}-{NNN}— domain UPPERCASE, numbers zero-padded to 3 digits - Sequential per sub-feature folder — each sub-feature starts at 001
- IDs are stable — never renumber when adding new stories
Story File Template
Each .md file contains exactly one story:
Rules:
- Actor is inline (e.g. "buyer", "seller", "admin") — no upfront persona definitions
- Keep "I want" focused on ONE capability — split if it does two things
- Write 2–5 acceptance criteria per story
- Cover the happy path first, then key edge cases
- Each criterion must be testable — no vague language like "should work well"
Domain Index Template
Each domain folder has an index.md:
Update the index whenever stories are added, removed, or renamed.
Workflow: Interview Mode
1. Understand the Input
Ask: "What are we writing stories for?"
- PRD or document provided → Read it fully, extract features, confirm understanding
- Raw feature spec → Parse it, ask clarifying questions
- Verbal description → Interview to build understanding
- Existing codebase → Explore the code, infer current behavior, identify gaps
2. Identify Domains and Sub-features
Propose domains (bounded areas of functionality), then sub-features within each:
Based on what you've described, I see these domains:- {domain-a} — {brief description}- {sub-feature-1}, {sub-feature-2}- {domain-b} — {brief description}- {sub-feature-1}, {sub-feature-2}Does this grouping make sense? Anything to add, merge, or split?
Also identify actors that interact with each domain.
3. Interview Per Sub-feature
For each sub-feature, drill into specifics. Ask 2–4 questions per round:
- "What's the happy path for {action}?"
- "What happens when {edge case}?"
- "Can a {actor} do {action} without {precondition}?"
- "What should the system do if {failure scenario}?"
- "Are there time-based rules? (expiry, cooldowns, windows)"
Resolve answers before moving on.
4. Draft and Approve Stories
Once you have enough understanding, draft the stories and present them:
Here are the draft stories for {domain} / {sub-feature}:US-{DOMAIN}-001: {Title}As a {actor}, I want {capability}, so that {benefit}.US-{DOMAIN}-002: {Title}As a {actor}, I want {capability}, so that {benefit}.Want me to show the full acceptance criteria, adjust these, or move on?
5. Write Files
After user approval, create the story files and update the domain index.md.
6. Summary
After all domains are done:
User stories created:| Domain | Sub-feature | Stories | Actors ||--------|-------------|---------|--------|| {domain-a} | {sub-feature} | US-{DOMAIN}-001 – 003 | {actor} || {domain-b} | {sub-feature} | US-{DOMAIN}-001 – 002 | {actor-a}, {actor-b} |
Updating Existing Stories
- Adding stories to existing sub-feature: Continue from the next number, update
index.md - Adding new sub-feature: Create folder, start at 001, add section to
index.md - Adding new domain: Create domain folder with
index.md, then sub-features as normal
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckinguser-story/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/raphaelmans/agent-skills/user-story)