planning — the walking skeleton
The walking skeleton — the value test
Release 1 — Slice 1 — is the walking skeleton. It is the smallest release that delivers, end-to-end, a coherent change in Gal's experience. Not a prototype. Not a partial. A real, shippable, end-to-end version of the new flow — just thinly. Gal can sit down on a Tuesday afternoon, see her 28 students, mark each one, correct mistakes, and submit. The MCQ scoring is automatic. The submission produces a confirmation. None of the scoring is rubric-based, none of the views are filterable, none of the edge cases are smoothed. But the flow works.
The walking skeleton is the discipline. It says: before we add depth to any column, we have a complete row. A complete row is what changes Gal's situation. A column without depth is still useful when paired with the other columns; a column built deep with no other columns doesn't change anything. A team that builds depth before width — that finishes the marking Epic before starting the submission Epic — has spent six weeks producing nothing Gal can use.
The walking skeleton is what closes the loop with the brief. It is the smallest version of the prediction we can actually test.
Why the walking skeleton replaces Must / Should / Could
Many teams use Must / Should / Could (MoSCoW) to decide what ships in a release. Must items are required; Should items ship if possible; Could items go to the next release. The framework is well-known and easy to apply.
It is also a quiet substitution. Must / Should / Could measures the team's commitment, not the value to the person. A release with all Must items done can ship — even when it doesn't change anything for Gal. The team has finished what it promised; the situation is unchanged. The framework optimises for the wrong thing, because it asks "what did we say we'd do" instead of "what changes the person's day."
The walking skeleton replaces this. It asks one question: does this slice, end-to-end, change the situation the brief named? If yes, the slice is the release. If no, the slice is not yet a release — it is a partial slice, and the team needs to add the cards that complete the row, or remove cards that aren't pulling weight. The map shows the answer visibly. There is no way to satisfy the walking-skeleton test by checking boxes.