Skip to content

SKILL.md

Latest commit

 

History

History
108 lines (80 loc) · 3.77 KB

SKILL.md

File metadata and controls

108 lines (80 loc) · 3.77 KB
name epic-breakdown-advisor
description Guide splitting large epics into right-sized user stories using Richard Lawrence's 9 proven splitting patterns. Recommends the best pattern based on epic characteristics.
type interactive
domain planning
difficulty intermediate
estimated_time 20-30 min
prerequisites
epic-hypothesis
user-story
outputs
split-stories
triggers
This epic is too big
Help me break down this epic
Split this into user stories
Epic breakdown

Purpose

Guide splitting large epics into right-sized user stories using proven patterns. Recommends the best splitting strategy based on the epic's characteristics — workflow steps, business rules, data variations, user personas, or platform differences.

Key Concepts

9 Splitting Patterns (Richard Lawrence)

# Pattern When to Use Example
1 Workflow Steps Epic follows a multi-step process "Checkout" → Cart, Payment, Confirmation
2 Business Rules Different rules for different cases "Apply discount" → By tier, volume, promo
3 Happy/Unhappy Path Error handling adds complexity "Login" → Success, Wrong password, Locked
4 Input Variations Multiple data types or sources "Import data" → CSV, API, manual entry
5 Data Types Different entities need same operation "Notifications" → Email, SMS, push
6 Operations (CRUD) Create, Read, Update, Delete "Manage users" → Create, view, edit, disable
7 Role/Persona Different users need different versions "Dashboard" → PM view, Exec view, Dev view
8 Platform Needs to work across contexts "Search" → Web, mobile, API
9 Performance Functional first, optimize later "Report" → Generate (slow), then optimize

Application

Question 1: Describe the Epic

Agent asks: "What's the epic? Share the title and a 2-3 sentence description."

Question 2: Epic Characteristics

Agent asks: "What makes this epic big? (Select all that apply)"

  1. Multiple workflow steps
  2. Complex business rules
  3. Multiple user types
  4. Multiple platforms/channels
  5. Error handling / edge cases
  6. Performance requirements
  7. Multiple data types

Question 3: Apply Patterns

Agent recommends 2-3 splitting patterns and generates stories:

## Recommended Split: [Pattern Name]

### Original Epic
[Epic description]

### Split Stories
1. **[Story 1]:** As a [persona], I want to [first slice], so that [value]
2. **[Story 2]:** As a [persona], I want to [second slice], so that [value]
3. **[Story 3]:** As a [persona], I want to [third slice], so that [value]
...

### Validation
- [ ] Each story delivers independent value
- [ ] Stories are small enough for 1 sprint
- [ ] Combined stories cover the original epic
- [ ] Stories are testable independently

Quality Rubric

Dimension 0 1 2
Pattern Fit Random split Pattern applied but forced Pattern matches epic nature
Story Independence Stories depend on each other Some independence Each story deployable alone
Coverage Stories miss parts of epic Most covered Full epic coverage + validation
Size Still too big (multi-sprint) Mostly right-sized All stories fit in 1 sprint

References

Related Skills

  • skills/specification/user-story/SKILL.md — Write the resulting stories
  • skills/specification/user-story-splitting/SKILL.md — Splitting patterns reference
  • skills/planning/epic-hypothesis/SKILL.md — Epic that needs splitting

External Frameworks

  • Richard Lawrence, Patterns for Splitting User Stories (AgileForAll)

Skill type: Interactive Domain: Planning & Prioritization