Epic Breakdown Advisor (Humanizing Work)
Break epics into user stories using the Humanizing Work methodology — INVEST validation, 9 splitting patterns applied sequentially, and evaluation criteria that expose low-value work.
Your epic is "Manage user profiles," estimated at 3 weeks, and the team is arguing about whether to split it by screen, by layer, or by user type. The Humanizing Work methodology answers that: INVEST first, then walk 9 patterns in order, then evaluate the split for whether it reveals work you can kill. The word "manage" alone tells you the pattern.
Who it's for: PMs splitting epics before sprint planning, scrum masters coaching story-splitting, tech leads who keep getting backlog items sized as 8-day monsters, agile coaches teaching vertical slicing, PM/eng pairs in backlog refinement
Example
"Help me split this onboarding epic — it's estimated at 12 days and we need to sprint it" → INVEST check + apply Pattern 1 (workflow) with thin end-to-end slices + each story covers full flow with increasing sophistication + split evaluation flags low-usage steps to deprioritize + final 4 stories at 2-3 days each
CLAUDE.md Template
New here? 3-minute setup guide → | Already set up? Copy the template below.
npx ccplaybooks add epic-breakdown-advisor# Epic Breakdown Advisor
Guide product managers through breaking down epics into user stories using Richard Lawrence's complete Humanizing Work methodology—a systematic, flowchart-driven approach that applies 9 splitting patterns sequentially. Each split preserves **vertical slices** (end-to-end value) rather than horizontal slices (technical layers).
This is not arbitrary slicing—it's a proven, methodical process: validate → apply patterns in order → evaluate results strategically.
## Three-Step Process
1. **Pre-Split Validation:** Check INVEST criteria (except "Small")
2. **Apply Splitting Patterns:** Work through 9 patterns sequentially until one fits
3. **Evaluate Splits:** Choose the split that reveals low-value work or produces equal-sized stories
Get new playbooks like this one
One email a week with new Claude Code workflows. Free, like everything here.
No spam. Unsubscribe anytime.
What This Does
Walks through Richard Lawrence's Humanizing Work story-splitting methodology: INVEST validation → apply 9 splitting patterns sequentially → evaluate splits for waste-elimination and size equality. Produces vertical slices (end-to-end value), not horizontal slices (technical layers).
Not arbitrary slicing. A systematic, flowchart-driven process.
Quick Start
mkdir -p ~/Documents/EpicBreakdown
mv ~/Downloads/CLAUDE.md ~/Documents/EpicBreakdown/
cd ~/Documents/EpicBreakdown
claude
Paste your epic (title, description, acceptance criteria, persona, rough estimate). Claude validates INVEST, applies patterns in order, and delivers the breakdown.
The 9 Patterns (Applied in Order)
| # | Pattern | Signal | Example |
|---|---|---|---|
| 1 | Workflow Steps | Multi-step process | Thin end-to-end slices, not step-by-step |
| 2 | Operations (CRUD) | Word "manage" | Create/Read/Update/Delete stories |
| 3 | Business Rule Variations | Different rules per segment | One story per rule |
| 4 | Data Variations | Different data types | Just-in-time additions |
| 5 | Data Entry Methods | Fancy UI | Simple input → sophisticated UI later |
| 6 | Major Effort | Infrastructure build | First implementation + trivial additions |
| 7 | Simple/Complex | Has optional complexity | Core simplest + variations |
| 8 | Defer Performance | Non-functional requirements | Make it work → make it fast |
| 9 | Break Out a Spike | High uncertainty | Time-box investigation, then restart |
Pre-Split INVEST Check
Before splitting, verify (except "Small"):
- Independent — No blocking dependencies
- Negotiable — Collaborative discovery, not prescriptive spec
- Valuable — Observable user value (if NO: STOP, reframe)
- Estimable — Team can size
- Testable — Concrete acceptance criteria
Split Evaluation (Step 3)
- Does this reveal low-value work to kill or defer? Good splits expose the 80/20.
- Are stories equally sized? Equal sizes give the PO prioritization flexibility.
If neither → try a different pattern.
Tips & Best Practices
- Vertical slices always. "Build API" + "Build UI" fails — neither delivers value. Each story touches all layers.
- Pattern 1 done right = thin end-to-end. Story 1 is full workflow with simplest path; Stories 2+ add sophistication.
- The word "manage" is diagnostic. It signals CRUD — split by operation.
- After splitting, re-check story sizes. If any > 5 days, restart Pattern 1 for that story.
- Cynefin matters. Complex domain = 1–2 learning stories, not exhaustive enumeration.
Common Pitfalls
- Skipping INVEST validation and splitting technical tasks as if they were stories
- Step-by-step splits that don't deliver end-to-end value
- Horizontal (layer-based) slicing that masquerades as vertical
- Forcing a pattern that doesn't fit instead of moving to the next
- Not re-splitting stories that are still too large after one pass