How to create a story bible for a novel.

1. Treat it as a reference, not a wiki. A bible only pays for itself if you — or something reading on your behalf — can query it mid-draft: "how old is she in this chapter," "what does this substance do here," "did the Order ever hold office." A page of running notes forces a reread to find one fact. A reference built for lookup gets you the answer in seconds.
2. Capture by kind of fact, not by chapter. Give characters, locations, cultures, the world's rules and systems, and its history and timeline each their own kind of entry — structured the same way every time: a name, a short description, the handful of facts that actually matter, and how it connects to everything else. One structured entry per fact, so the bible stays a set of lookups instead of one long document.
3. Start from a premise, or start from a draft. Still planning? Sketch the spine first — the handful of characters, the one rule of magic or trade, the one faction that will carry the plot — and let entries accumulate as you outline. Already have pages written? Work backwards: read the draft and mine it for what it has already established, one fact at a time, instead of inventing canon that quietly contradicts what's already on the page.
4. Record where each fact came from. Every entry should cite the passage that put it into canon — the chapter, ideally the line. That citation is what keeps the bible anchored to the book instead of drifting into a parallel document with its own opinions. When a fact needs to change, you can find every passage it touches instead of guessing.
5. Keep it lean and living. Add only what the prose needs. A bible padded with lore the book never uses is nearly as useless as no bible at all — it takes just as long to search and rewards you less. Update entries as the manuscript grows rather than saving one big pass for the end; a bible that's a chapter behind the draft is already wrong.

# Character — <name>

Role in the story:

## Established
- Green-eyed, and says so herself — ch. 3, "The Hall of Tides"

## Planned
- Learns her mother lied about the crossing

# Relationship — <A> and <B>

Nature:

## Established
- They have not spoken since the funeral — ch. 1

## Planned
- She learns he was right about Hendricks, and cannot take back the year of not believing him

# Location — <place>

What it is:

## Established
- Two days' ride south of the capital — ch. 5
- Held by the harbour guild — ch. 5

## Planned
- Burns in the third act

# Timeline — <event>

What it changes:

## Established
- The crossing happened eleven years before ch. 1 — ch. 2

## Planned
- The second crossing, timing not yet fixed

# World rule — <rule>

The rule:
What it costs:
What it cannot do:

## Established
- Calling the tide takes a day of the caller's life — ch. 4

## Planned
- Whether the cost can be paid by someone else

# Object — <thing>

What it is:

## Established
- Passed to Tarsil at the docks — ch. 6

## Planned
- Who ends up holding it, still undecided

# Knowledge state — <fact>

The fact:

## Established
- Tarsil knows; Neya does not — ch. 6

## Planned
- Neya learns it from the harbour guild rather than from Tarsil

# Unresolved thread — <question>

The question the reader is holding:

## Established
- Raised, unanswered — ch. 2

## Planned
- Answered in the third act, or deliberately left open

For a standalone novel, one file of these entries is usually enough. For a multi-book series, keep world rules, locations and knowledge states in one shared set and give each volume its own characters, timeline and threads — the shared half is what has to stay true across books, and the per-volume half is what changes. On planning that shape deliberately, see how to plan a book series and writing a series in Loreon.

• The wiki nobody rereads. A bible built once and never queried is just a longer document to forget. If nothing — you or a tool — reads it back against the prose, it isn't earning its place.
• Over-detailing lore the book never uses. Three pages on a dynasty's succession law that the plot mentions once is effort spent where the reader will never look. Depth belongs where the story actually goes.
• Letting the bible drift from the manuscript. A revision changes a name, an age, a rule — and the entry doesn't get updated. From then on the bible is actively wrong, which is worse than having no bible.
• One long document instead of structured entries. A single sprawling file with everything in it can't be queried, can't be checked against automatically, and gets harder to search every chapter you add to it.

• When one fact lives in several entries. A character's age, a journey's length, the year of a war — each is written down in more than one place, and changing it means finding every copy. The template cannot tell you where the copies are. This is the first thing to bite, and it bites on revision rather than on drafting.
• When several characters know different things at different times. Knowledge states are the hardest entry to keep by hand, because the answer to "who knows this in chapter 14?" depends on every scene before it. One or two secrets are tractable. A dozen, across forty chapters, is bookkeeping you will stop doing.
• When continuity has to hold across volumes. Within one book you can reread to check. Across three, the thing you need to verify is in a manuscript you finished two years ago, and rereading it is no longer a check you will actually perform.

The common thread is not size — it is cross-referencing. A template stores facts; what it cannot do is tell you every place a fact is relied on. When that is the work you are spending your time on, software starts being worth it, and a bible the software reads is the shape that helps.