Epic Breakdown Advisor - An Interactive Advisor for Breaking Down Epics into User Stories

Skill Overview

Epic Breakdown Advisor is an interactive breakdown advisor based on the Humanizing Work methodology. Through an INVEST pre-check and nine splitting patterns, it systematically breaks down epics that are too large or difficult to estimate into Sprint-sized user stories while preserving end-to-end user value.

Applicable Scenarios

1. Breaking down large requirements before Sprint planning: An epic in the backlog is too large, and the team cannot estimate or prioritize it. It needs to be broken down into user stories that can be completed and safely delivered within 1–5 days.
2. Turning vague feature blocks into stories: A one-line large requirement such as “Support exporting reports to Excel/PDF/CSV with customizable date ranges” needs to be converted into specific, implementable, and verifiable stories.
3. Breaking down multi-step workflows and feature variants: Epics involving approval processes, multiple business rules, various data types, or “management” CRUD operations require the right splitting pattern before creating vertical slices.

Core Features

1. INVEST pre-breakdown validation: Before splitting, check independence, negotiability, value, estimability, and testability one by one to avoid forcing technical tasks into user stories.
2. Step-by-step guidance through nine splitting patterns: Match patterns in sequence—workflow steps (thin end-to-end slices), CRUD operations, business rule variants, data variants, data entry methods, primary effort, simple/complex, latency and performance, and extracting a Spike. Through a question-and-answer process, the advisor helps select the right pattern and generate specific stories.
3. Breakdown quality assessment and deliverables: Evaluate the breakdown from two perspectives: whether it exposes low-value work that can be cut, and whether it produces stories of comparable size. Finally, generate a complete breakdown plan with Given/When/Then acceptance criteria.

Frequently Asked Questions

What should we do if an epic is too large for the team to estimate?

Start with an INVEST pre-check to confirm that it is truly a user story delivering user value. Then try the nine splitting patterns in sequence. If none of the first eight patterns applies, the uncertainty is likely too high. The advisor will recommend first conducting a 1–2 day, time-boxed Spike investigation, then returning to split the epic once the questions have been answered.

Why should we preserve vertical slices instead of splitting by frontend and backend layers?

Horizontal slices—such as “build the backend API first, then the frontend page”—cannot independently deliver user-perceivable value. Vertical slices span multiple architectural layers, allowing each story to deliver an observable change in system behavior. This enables the team to deliver and validate stories incrementally while adjusting priorities at any time.

What should we do if the resulting stories are still too large?

The skill output provides a clear next step: for any story that exceeds five days, start again with Pattern 1 and split it another round until all stories fall within the 1–5-day range. At the same time, check whether the breakdown has exposed low-value work that can be removed or deferred, reducing unnecessary development effort at the source.