StoryGraph

The knowledge layer — who knows what

A story isn’t just what happens — it’s who knows it, and when. The knowledge layer is the part of the StoryGraph that tracks exactly that: what the reader knows versus what each character knows, at every beat, across the whole book. It’s what makes a mystery hold together and a plot hole impossible to write by accident.

Why it’s the hard part. A prose-first tool has to remember who knows what across a hundred pages — and it forgets, so characters suddenly know things they were never told. The knowledge layer makes it explicit and checkable, so the story can withhold, reveal and pay off information on purpose.

The four moves

Every beat can make four kinds of move on what’s known. You author them in the beat’s knowledge layer:

Establishes (for all)

public facts — known to everyone, including the reader

The ground truth a beat sets down. Once established, everyone — every character and the reader — knows it, and the writing must stay consistent with it.

Reveals (secret → to whom)

a secret disclosed to specific observers (reader★ = the reader)

A fact let out to some observers but not others. This is what makes mystery and dramatic irony possible — the reader can know something a character doesn’t, or the reverse.

Requires (precondition)

the character must already know these facts — otherwise a plot hole

What a beat assumes is already known. If a required fact hasn’t been established or revealed by this point, the engine flags a plot hole — before a word is written.

Overturns

which prior beliefs this beat nullifies (a turn)

The belief a beat cancels. When a twist lands or an arc turns, the old understanding is retired so it stops colouring what comes next.

The reader is a character too

The trick that makes it all work: the reader is tracked as an observer just like any character (marked reader★). So you can reveal a fact to the reader but not the hero — dramatic irony — or to the hero but not the reader — a secret the detective keeps until the last page. The graph knows the difference, and the writing respects it.

Facts that expire

A truth doesn’t always hold for the whole book. Two markers give a fact a validity window: one makes it take effect only from a chosen scene onward (so canon can’t reveal an event before the prose has shown it), and one retires it after a chosen scene (so a Part-1-only truth doesn’t bleed into Part 2). It’s how a story can honestly change — “the hero is an ordinary soldier” can be true early and stop being true later, without contradiction.

The payoff. Because the graph tracks knowledge instead of hoping the model remembers it, it can do what prose-first tools can’t: withhold a fact until exactly the right beat, guarantee a character never acts on something they weren’t told, and pay off a planted secret precisely on time — across a whole book, not just a page.