User Story Generation

User Story Generation Skill

Transform raw requirements into actionable, INVEST-compliant user stories ready for sprint planning.

Core Process

Step 1: Analyze Input

Read and understand the provided content:

• Meeting transcripts or notes
• Feature requests or requirements documents
• Stakeholder conversations
• Product vision statements

Step 2: Identify Pain Points

Extract both explicit and implicit pain points:

• Explicit: Directly stated problems in the text
• Implicit: Inferred needs (e.g., request for automation implies manual process is tedious)

Consult references/invest_and_splitting.md for pain point analysis techniques.

Step 3: Generate User Stories

Create stories using the format:

As a [type of user],
I want [goal/desire],
given [context/constraints],
so that [benefit/value].

Requirements:

• Each story must have a clear pain point it addresses
• Stories must be Independent, Negotiable, Valuable, Estimable, Small, and Testable (INVEST)
• Keep stories small enough to complete in 1-3 days
• Stories should be vertically sliced (not technical layers)

Acceptance Criteria: For each story, generate 2-4 testable acceptance criteria using Given/When/Then format:

- Given [initial context], when [action occurs], then [expected outcome]
- Given [initial context], when [action occurs], then [expected outcome]

Acceptance criteria should:

• Be specific and testable
• Cover the main success scenarios
• Include edge cases where relevant
• Be clear enough for QA to validate

Consult references/invest_and_splitting.md for INVEST criteria details and story-splitting techniques when stories are too large.

Step 4: Organize Stories

Group related stories into logical categories/epics:

• By feature area (e.g., "User Authentication", "Payment Processing")
• By user journey (e.g., "Onboarding", "Checkout Flow")
• By system component (e.g., "Admin Dashboard", "Mobile App")

Step 5: Ask for Output Format

Always ask the user which output format they prefer:

• CSV: Best for importing to Excel/Google Sheets
• Markdown Table: Best for documentation, Notion, GitHub
• JSON: Best for programmatic processing or API integration
• Excel (.xlsx): Best for formatted spreadsheets with styling

Do NOT assume a format - always ask first.

Step 6: Generate Output

Required Columns

Every story must include these seven fields:

1. category_epic: High-level grouping for related stories
2. title: Brief summary (5-10 words)
3. user_story: Complete story in "As a... I want... given... so that..." format
4. acceptance_criteria: 2-4 testable criteria in Given/When/Then format (bullet points)
5. requirement: Original requirement or feature description from input
6. pain_point: The specific problem this story solves
7. created_date: Today's date in YYYY-MM-DD format

CSV Format

Category/Epic,Title,User Story,Acceptance Criteria,Requirement,Pain Point,Created Date
User Authentication,Login with Email,"As a returning user, I want to log in with my email and password, given I have an existing account, so that I can access my personalized content","- Given valid credentials, when I submit the login form, then I am redirected to my dashboard
- Given invalid credentials, when I submit the login form, then I see an error message
- Given I am already logged in, when I navigate to the login page, then I am redirected to my dashboard","Users need to access their accounts","Users cannot access their saved preferences",2026-01-29

Markdown Table Format

| Category/Epic | Title | User Story | Acceptance Criteria | Requirement | Pain Point | Created Date |
|---------------|-------|------------|---------------------|-------------|------------|--------------|
| User Authentication | Login with Email | As a returning user... | - Given valid credentials, when I submit the login form, then I am redirected to my dashboard- Given invalid credentials, when I submit the login form, then I see an error message | Users need to... | Users cannot... | 2026-01-29 |

JSON Format

[
  {
    "category_epic": "User Authentication",
    "title": "Login with Email",
    "user_story": "As a returning user, I want to log in with my email and password, given I have an existing account, so that I can access my personalized content",
    "acceptance_criteria": "- Given valid credentials, when I submit the login form, then I am redirected to my dashboard\n- Given invalid credentials, when I submit the login form, then I see an error message\n- Given I am already logged in, when I navigate to the login page, then I am redirected to my dashboard",
    "requirement": "Users need to access their accounts",
    "pain_point": "Users cannot access their saved preferences",
    "created_date": "2026-01-29"
  }
]

Excel Format

For Excel output:

1. Generate the JSON format first
2. Save JSON to a temporary file (e.g., /home/claude/stories.json)
3. Run the Excel generation script with today's date in filename:

``bash cd /home/claude/user-story-generation/scripts python3 generate_excel.py --input /home/claude/stories.json --output /home/claude/$(date +%Y-%m-%d)_userstories.xlsx ``

1. Move the generated file to outputs:

``bash mv /home/claude/*_userstories.xlsx /mnt/user-data/outputs/ ``

1. Present the Excel file to the user

The script automatically applies formatting:

• Colored header row (blue background, white text)
• Frozen header for scrolling
• Wrapped text for readability
• Optimized column widths
• Professional borders
• Seven columns: Category/Epic, Title, User Story, Acceptance Criteria, Requirement, Pain Point, Created Date

Quality Guidelines

Story Title:

• Keep to 5-10 words
• Descriptive and actionable
• Avoid technical jargon

User Story:

• Always use full "As a... I want... given... so that..." format
• Be specific about user type (admin, customer, guest, etc.)
• Include meaningful context in "given" clause
• Articulate clear value in "so that" clause

Acceptance Criteria:

• Write 2-4 criteria per story in Given/When/Then format
• Each criterion should be independently testable
• Cover main success path and key edge cases
• Be specific about expected outcomes
• Use bullet points for readability
• Example format: "- Given [context], when [action], then [outcome]"

Requirement:

• Extract the original requirement from input text
• Keep concise but informative
• Preserve key details from original source

Pain Point:

• State the specific problem being solved
• Connect to user or business value
• One clear pain point per story

Common Patterns

Authentication Stories:

• Category: "User Authentication" or "Security"
• Pain points: Access control, security concerns, user experience

Data Management Stories:

• Category: "Data Management" or "[Entity] Management"
• Pain points: Manual processes, data accuracy, efficiency

Reporting/Analytics Stories:

• Category: "Reporting" or "Analytics"
• Pain points: Lack of visibility, manual reporting, decision-making delays

Integration Stories:

• Category: "Integrations" or "Third-Party Services"
• Pain points: Manual data transfer, system disconnects, workflow inefficiency

Resources

• INVEST Criteria & Story Splitting: references/invest_and_splitting.md
• Excel Generation Script: scripts/generate_excel.py

Load the INVEST reference when you need guidance on:

• Validating story quality
• Splitting large stories
• Understanding pain point analysis

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Luisarueda1
• Source: Luisarueda1/user-story-generation-skill
• License: MIT

Install and usage instructions live in the source repository linked above.