Domain Story Command

Create domain stories using pictographic storytelling methodology to capture business workflows.

Arguments

• <journey-name>: Name of the journey/process to capture (e.g., "order-fulfillment", "customer-onboarding")

• --type: Story type (default: as-is)

  • as-is: Document current state (how things work today)
  • to-be: Document desired future state (how things should work)

• --dir: Output directory (default: docs/domain-stories/)

Execution

Step 1: Parse Arguments

Extract journey name and type from arguments.

If no journey name provided, ask the user: "What business process or journey would you like to capture as a domain story?"

Default type is as-is unless specified.

Step 2: Load Domain Storytelling Skill

Invoke the domain-storytelling skill to access pictographic notation, story templates, and boundary discovery patterns.

Step 3: Spawn Story Facilitator

Launch the story-facilitator agent to guide the interactive storytelling session.

The agent will:

1. Collect the Story

   • Ask the user to describe the process in their own words
   • Listen for actors, work objects, and activities
   • Capture the sequence of steps

2. Refine the Story

   • Explore edge cases: "What happens when X fails?"
   • Identify variations: "Are there different paths?"
   • Note exceptions: "What if Y isn't available?"

3. Identify Actors

   • Human actors (roles, not individuals)
   • System actors (internal and external systems)
   • Catalog with responsibilities

4. Catalog Work Objects

   • Documents that flow through the process
   • Data entities that are read/written
   • Physical items (if applicable)

5. Structure the Story

   • Number each activity in sequence
   • Create the story diagram
   • Note terminology and language patterns

Step 4: Generate Output Artifacts

Story Diagram

# Domain Story: [Journey Name]

**Type:** AS-IS | TO-BE
**Scope:** [Starting Point] → [Ending Point]
**Date:** [ISO Date]
**Captured By:** AI-assisted session

## Story Sequence

| # | Actor | Activity | Work Object | Recipient |
|---|-------|----------|-------------|-----------|
| 1 | Customer | submits | Order Form | Sales Rep |
| 2 | Sales Rep | checks | Product Catalog | - |
| 3 | Sales Rep | creates | Quote | - |
| 4 | Sales Rep | sends | Quote | Customer |
| 5 | Customer | approves | Quote | Sales Rep |
| 6 | Sales Rep | creates | Order | Order System |

## Visual Representation

[Mermaid sequence diagram if helpful]

## Actors

| Actor | Type | Responsibilities |
|-------|------|------------------|
| Customer | Human | Initiates orders, approves quotes |
| Sales Rep | Human | Processes orders, creates quotes |
| Order System | System | Stores and tracks orders |

## Work Objects

| Work Object | Owner | Lifecycle |
|-------------|-------|-----------|
| Order Form | Customer | Created → Submitted → Archived |
| Quote | Sales Rep | Created → Sent → Approved/Rejected |
| Order | Order System | Created → Fulfilled → Closed |

## Glossary

| Term | Definition | Context |
|------|------------|---------|
| Quote | Price estimate for customer | Sales process |
| Order | Confirmed purchase request | Fulfillment |

## Boundary Observations

Potential bounded context boundaries observed:
- **Sales Context**: Quote creation, customer interaction
- **Order Context**: Order processing, fulfillment tracking

Language shifts noted:
- "Quote" becomes "Order" after approval
- "Customer request" becomes "work item" in fulfillment

Step 5: Suggest Next Steps

After completing the story, suggest:

1. If AS-IS captured: "Would you like to create a TO-BE story showing the desired future state?"

2. Related stories: "Are there related processes that connect to this one?"

3. Event storming: "This story can feed into an event storming session to design the domain model. Run /event-storm [domain] when ready."

4. Architecture documentation: "The bounded context observations can inform architecture decisions."

Step 6: Save Results

If the user wants to save results, write the story to:

• docs/domain-stories/[journey-name]-[type]-[date].md Use --dir for custom location.
• Or ask user for preferred location