/expand — Use Case Expansion

Transform high-level epics into comprehensive use case specifications by uncovering
explicit, implicit, and cascading use cases.

Step 0 — Parse Arguments

From the invocation args, extract:

• epic: File path to epic document OR inline epic text
• --output : Directory to write output files (default: stdout)
• --concise: Final results only, skip discovery reasoning

Validate:

• If file path: verify it exists with Read tool
• If --output: verify directory exists
• If neither file nor text: ask for the epic

Step 1 — Triage

Estimate epic complexity:

Fast path (inline expansion):

• Short epic (< 500 words)
• Single user role, clear scope
• Proceed to Step 2a

Agent path (full expansion):

• Long epic or multi-role system
• Complex domain with regulatory/compliance needs
• Multiple integration points
• Proceed to Step 2b

Step 2a — Inline Expansion

For the epic, discover use cases in three layers:

1. Explicit: Directly stated requirements and user actions
2. Implicit: Unstated but necessary (auth, error handling, data validation, empty states)
3. Cascading: Second-order effects (what happens when X fails? what if concurrent users?)

Format each use case:

### UC-[N]: [Name]
**Actor**: [who]
**Trigger**: [what starts this]
**Flow**: [numbered steps]
**Alternatives**: [error paths, edge cases]

Step 2b — Agent Dispatch

Use the Agent tool:
  subagent_type: "use-case-expander"
  prompt: "Expand this epic into comprehensive use cases: [epic content].
           [If --output: write results to [dir]]
           [If --concise: final results only, skip discovery reasoning]"

Step 3 — Output

After expansion:

• If --output: write to specified directory
• If --concise: present only the final use case list
• Otherwise: include discovery reasoning and the use case list
• Summarize: "[N] use cases discovered ([X] explicit, [Y] implicit, [Z] cascading)"