The short answer
A series stays consistent when every book reuses the same facts about people, places, rules, and dates. Memory alone cannot carry that load past book two, so you keep one file where the facts live and check it before you draft. Enter your planned book count, viewpoint count, a one-line setting, a one-sentence rule, and how many books are already written, and this starter hands back a bible skeleton with five named sections, exact entry counts, a per-book tracking table, and a checklist ordered by what to fill first. You supply every fact yourself. The tool only decides how many slots, rows, and fields your plan requires, using fixed arithmetic that gives the same answer for the same inputs every time.
What goes inside a series bible
A series bible is a single working file that records the facts later books must not contradict. It holds five kinds of information, and each kind earns its own section. Characters covers who people are and what they want. Places covers where events happen and who controls each location. Timeline covers when things happened and who knows about them. Rules covers what your magic or technology can and cannot do. Open threads covers the promises you have made readers that no book has paid off yet.
A bible is not an outline, not a draft, and not a place for prose. It stores short declarative facts, one per field, so a tired author at midnight can verify a detail in seconds. Brevity is the point, because a reference file you avoid reading helps nothing.
Size the bible to the series you are actually writing. A standalone with one viewpoint needs a handful of pages. A six-book epic with five viewpoints needs a far larger file, and starting too small means rebuilding it mid-series when contradictions are already printed. The starter on this page computes that size for you before you type a single fact.
How this starter sizes your bible
The tool reads five inputs and applies fixed rules to each one. Planned books set the timeline rows at three per book and the thread rows at two per book, because every installment adds dated events and new promises. Viewpoint count sets the character slots at two per viewpoint plus two for the wider cast, since each viewpoint needs a lead entry plus a close supporter, with two recurring minor figures shared across the file. Books already written raise the contradiction-risk score, because printed facts cannot be revised away. Your setting line seeds the Places section, which always opens five location slots. Your rule sentence sets the rule depth: ten words or fewer counts as plain and needs four fields, eleven to twenty words counts as moderate and needs six, and anything longer counts as layered and needs eight.
Every field count then feeds one total. Each character slot carries five fields, each place carries four, each timeline row carries three, the rule section carries its depth number, and each thread row carries four. The risk score adds two points per viewpoint, up to four points for rule depth, up to four points for books already written, and two more points when the plan reaches five books or beyond. Scores of five or below read as low risk, six through ten read as caution, and eleven or above read as high risk that demands tending from book one.
Check the default sizing with your own arithmetic. Three planned books give nine timeline rows and six thread rows. Two viewpoints give six character slots. One written book contributes one risk point. The default rule sentence runs thirteen words, so it counts as moderate and needs six fields with two risk points. The field total is thirty for characters, twenty for places, twenty-seven for timeline, six for rules, and twenty-four for threads, which sums to one hundred seven. The risk total is four for viewpoints, two for the rule, one for the written book, and zero for length, which sums to seven, a caution reading. Same inputs always return these same numbers.
Illustrative hypothetical example: the blockade-city trilogy
Consider an illustrative hypothetical author planning three books around a coastal trading city under blockade, with two viewpoints and one finished draft. Her rule sentence states that harbor signal-fire magic burns the keeper memory each time it is lit, which at thirteen words counts as moderate. The starter returns six character slots, five place slots, nine timeline rows, six rule fields, and six thread rows, for one hundred seven fields at a caution score of seven.
Her first job is backfilling book one before outlining book two. She records both viewpoint leads with a goal sentence each, adds two supporters, and adds two recurring minor figures: a harbormaster who controls the signal tower and a smuggler who carries messages through the blockade. She fixes nine dated rows covering the blockade declaration, the first signal firing, and the smuggler introduction, noting each time which characters witnessed what. She writes the rule once with its cost stated plainly, then adds the limit she actually used in the draft, which is that a keeper who fires twice in one night forgets the whole previous week. Finally she opens her thread ledger with the promises book one made: the missing blockade-runner brother, the foreign fleet on the horizon, and the keeper debt ledger. This example is hypothetical, and its names are invented to show the filling order, not to prescribe any story.
Illustrative hypothetical example: the six-book war epic
Now consider an illustrative hypothetical author with a harder problem: six planned books, five viewpoints, three drafts already written, and a rule sentence running twenty-five words across guild lodges, competing cost charts, generational debts, and storm visions. The starter returns twelve character slots, five place slots, eighteen timeline rows, eight rule fields, and twelve thread rows, for one hundred ninety fields at a score of nineteen, a high-risk reading.
His filling order differs because printed books constrain him. He cannot simplify the rule; three published volumes already show keepers paying different costs in different lodges. So his Rules pages must record each lodge variant as a separate stated fact with its own cost and limit, plus which books already show which variant, so future scenes stay inside established precedent. His eighteen timeline rows matter more than his character slots, because with five viewpoints the dangerous question is always who knew what and when. He backfills all three written books first, dating every lodge ruling and every fleet movement, before he outlines book four. He also caps his thread ledger hard: twelve rows exist, and each new book must close or carry one thread for every two it opens, or the promises outrun the payoffs. This example is hypothetical as well, drawn to show how scale changes the work rather than to describe any real series.
Characters, slot by slot
Give each viewpoint character a full five-field entry: name and role, one goal sentence, one wound or fear shaping their choices, two or three voice markers such as favored words or speech rhythms, and a per-book status line you update every volume. The goal sentence decides plots. The wound decides mistakes. The voice markers keep dialogue attributable when scenes grow crowded. The status line records where the character stands at each book end, so the next opening inherits position instead of guessing.
Supporting slots get the same five fields but shorter answers. A supporter needs a goal that intersects the lead, one distinguishing habit readers can picture, and a status line. Minor recurring figures need even less: name, function, one visual tag, and the thread they carry. Do not give every walk-on a full entry. The slots the starter counts already include room for the wider cast, so spend full entries on people who return and brief lines on people who appear once.
Record departures as facts, not intentions. When a character dies, leaves, or changes sides, update the status line with the book number and keep the old entries intact. Old entries explain earlier books. New status lines steer later ones. A bible that only shows the present cannot explain the past, and sequels constantly need the past.
Places that stay put
Places contradict series in quiet ways: a street changes direction, a harbor gains a second lighthouse, a travel time shrinks because the plot hurries. Fix five locations with four facts each: what the place looks like in two lines, what function it serves in the story, who controls it, and what changes about it across the books. Control matters most, because sieges, ownership disputes, and guild authority all turn on who holds which ground.
Break your one-line setting into anchors first. A coastal trading city under blockade already names water, trade, walls or patrols, and a controller enforcing the blockade. Each anchor becomes a place page or part of one. Then add the interiors where scenes actually happen: the signal tower, the market, the keeper lodge, the smuggler cove. Exteriors set scale. Interiors hold scenes. Both need the four facts.
Log travel times as place facts. Write how long the road from the north pass to the market takes in fair weather, and who can shorten it. Write how often ships run the blockade and what the crossing costs. Later books will compress journeys under deadline pressure, and a written time lets you notice the compression before readers do.
The timeline is a knowledge ledger
New authors treat timelines as date lists, but dates are only half the job. Each row needs the event, its date or season, and who knows about it. Knowledge tracking prevents the most common sequel error: a character acting in book three on information they never received in book two. When you record an event, name the witnesses in the same row.
Three rows per book is a floor, not a ceiling. Mark the political event that changes the board, the personal event that changes the lead, and the rule event that demonstrates what the magic or technology can do. If a book carries more turning points, add rows freely. The starter sizes the minimum so short books still get coverage; long books earn extra rows without permission.
Date relatively when your world lacks calendars: seasons, festivals, or reigns all work, provided every row uses the same scheme. Pick one frame in book one and never add a second without converting every old row.
Rules: write the cost before the wonder
Every rule page starts with the rule stated once, exactly as it works, in one or two sentences. Then fix the cost, the limit, and two edge cases you have already shown or plan to show. Plain rules stop there at four fields. Moderate rules add two more: what happens when the rule is pushed past its limit, and who teaches or enforces it. Layered rules add two further fields: known variants across lodges, schools, or models, and which books already demonstrate each variant.
Costs must be payable on the page. A cost the reader never sees a character pay reads as decoration, and decoration invites violation, because nothing reminds you it exists. If signal fire burns memory, show a keeper forgetting a name that mattered. If a gate closes at dusk, strand someone sympathetic outside it. Paid costs defend themselves. Unpaid costs evaporate by book two.
Edge cases deserve writing before you need them. Decide now whether the rule works on children, on the dying, on groups, at distance, through intermediaries, under coercion. You need not show every answer, but recording them stops a deadline scene from inventing an exception the whole series then inherits. The checklist reminds you to re-read this page before drafting any rule scene, and that single habit prevents more contradictions than any other in this guide.
Open threads: promises with due dates
A thread entry carries four facts: what the thread is, which book opened it, what promise it made the reader, and which book must pay it off. The promise field is the one authors skip and readers remember. A missing brother is not a thread; the promise that the search will resolve, and by when, is the thread. Write the promise in plain words: the brother will be found or confirmed lost by the end of book two.
Open about two threads per book and resolve or formally carry one each volume. This pace keeps forward motion without stacking unpaid debts. A carried thread needs a fresh on-page acknowledgment, a scene showing it still matters, so readers experience patience rather than neglect. Threads that vanish for two full books read as abandoned whatever your private plan says.
Close threads on the ledger when they pay off, with the book number and a one-line resolution. Paid rows are not clutter. They prove to your future self which promises are discharged, which stops book four from accidentally reopening a question book two already answered. A ledger that only lists the open items cannot distinguish resolution from forgetfulness.
What to fill in first after book one
Start with names and dates, because every other section references them. List every named character from the finished draft, then every dated or seasonal event, then the rule exactly as the draft shows it rather than as you intended it. Intention belongs in revision. The bible records what the manuscript actually says, since that is what readers of book one now believe.
Backfill before outlining forward. Read the finished book with the skeleton beside you and capture facts as you encounter them, resisting the urge to improve them mid-pass. Improvements are revision decisions for the draft, not silent edits to the reference file. If the draft says the crossing takes three days, the bible says three days, and any change goes through the manuscript first.
Then write the setting anchors and the thread promises. Anchors come straight from your one-line description, expanded into the five place pages. Promises come from rereading the final chapters and listing every question a reader would expect answered later.
Edge cases and failure modes
A bible fails in predictable ways, and each has a prevention. The commonest failure is the private-update problem: you change a fact in the draft and forget the bible, so the reference file becomes fiction and the draft becomes truth. Prevent it by updating both in the same sitting, draft first and bible second, before closing the file for the day.
The second failure is co-author drift, where two writers hold different versions of one fact. Keep one file with one approving owner; collaborators propose, the owner records.
The third failure is overbuilding before book one exists. A two-hundred-row timeline for an unwritten series encodes guesses, and guesses recorded as facts calcify into constraints that strangle the draft. The starter deliberately sizes minimums, not maximums. Fill the minimum from real manuscript evidence, then grow only where books demand it.
The fourth failure is the retcon trap in long series, where a new idea requires changing a printed fact. When that happens, record the change as a new dated entry with its reason, and note every earlier row it touches. Silent edits to old rows destroy the audit trail. Visible amendments preserve it, and they also force you to count the cost of the change before committing.
The fifth failure is format rot: the bible scatters across notebooks, margin comments, chat logs, and memory. Consolidate on a schedule, ideally at each draft completion, sweeping stray facts into the five sections and deleting the strays. A bible with one home gets consulted. A bible with five homes gets abandoned.
Edge cases deserve brief rulings. Prequels inherit the bible but add clearly marked earlier-dated rows; never renumber existing rows. Spin-offs copy only the sections they share and start their own thread ledger. Name collisions across books get distinguishing tags recorded once, such as elder and younger, with the tag used everywhere thereafter. Translations and adaptations never edit the master bible; they annotate copies. Each ruling takes one line to write and saves one contradiction to fix.
Genre guidance
Fantasy series live or die on rule pages and naming discipline. Invented names collide easily, so keep a running name list with one-line meanings or sounds-alike warnings, and check every new name against it. Rule variants across schools or bloodlines need separate entries with their own costs, because readers track which lineage pays what. Prophecies belong in the thread ledger with their exact wording, since paraphrase drift across books is a classic failure.
Science fiction series stress places and rule limits. Ships, stations, and colonies each need control, capacity, and travel-time facts, and technology needs stated failure modes, not just capabilities. Record what the drive, the network, or the weapon cannot do, and under which conditions it fails. Sequels escalate pressure by pushing systems toward those limits, which only works when the limits were fixed early.
Romance series center on character status lines and knowledge rows. Each book pairs new leads while earlier couples recur, so per-book status must record who is together, who knows about it, and what unresolved hurt each carries. Series-long emotional threads, such as a family rift behind the romances, belong in the thread ledger with a named payoff book, or they dissolve into background.
Mystery and thriller series hinge on the knowledge ledger most of all. Every clue, lie, and reveal needs a row naming who learned what and when, across both the standalone case and the series arc. The recurring detective needs a status line tracking injuries, reputation, and relationships that carry consequences between cases. Never let a sequel use case knowledge the detective never earned on the page.
Historical series add a research layer. Keep two timelines: the documented record with one-line sources, and your fictional events keyed to it. Deviations get flagged as deliberate with their reason, so book-four you can distinguish a chosen change from a remembered error. Period titles and terms need a small usage list, checked before each draft.
Keeping the bible alive across books
A bible dies when updating it feels like a separate project. Attach the update to a step you already perform: when a draft is declared finished, spend one session sweeping new names, dates, rule uses, and promises into the five sections. Mark each swept book with its completion date on the tracking table. The habit costs an evening per book and repays itself every drafting day after.
Review the whole file once per volume before outlining the next. Read characters for stale goals, places for changed control, and threads for overdue promises. This review often produces the next outline on its own, because overdue promises and changed control are plots waiting for a plan.
Prune with care as the series grows. Merge duplicate minor entries and archive resolved threads with payoff notes. Facts never change without a manuscript change behind them, so every edit traces to a page readers can check.
Questions authors ask next
Many authors wonder whether a short series truly needs this apparatus. The honest test is viewpoint count plus rule depth rather than book count. One viewpoint with a plain rule can often run on a few pages of notes. Two viewpoints with a moderate rule already generate enough names, dates, and costs to exceed working memory, which is exactly the default sizing on this page. Start the file when the plan crosses that line, even at two books.
Another frequent worry concerns timing: whether to build the file before drafting book one. Draft first, then extract. Pre-draft bibles record hopes, while post-draft bibles record evidence, and only evidence constrains later books correctly. Use the starter to learn the shape of the file, draft the book, then fill the skeleton from what the manuscript proves.
Authors also ask how strict the thread pace must be. Treat two opened per book with one closed or carried as a planning norm, not a law. Readers tolerate debt with a visible repayment schedule. They punish silent accumulation.
A further question is what to do when a contradiction is already printed. Record it, do not hide it. Add a dated amendment naming the conflict, choose the version future books will follow, and where possible write the reconciliation into the next story as a character correcting a record or a lodge admitting its chart was wrong. Readers forgive in-story corrections far more readily than silent edits, which they always notice.
Finally, authors ask how long filling the skeleton takes. The default size of about one hundred seven fields fills in two or three evenings from a finished draft, working section by section. Larger files scale roughly with the field total, so a one-hundred-ninety-field epic may take a full week of backfill. That week is the cheapest in the series, because every later contradiction costs more.
Where the actual writing happens
This page gives you the plan: sections, counts, tables, and order of work. The drafting itself belongs in a writing workspace that holds the manuscript alongside your notes. Do that work in Co-Writer at the writing desk, where each chapter draft can stay beside the bible facts it must honor. Open the skeleton this tool produced, keep it visible while scenes take shape, and let the recorded names, dates, and costs steady every new page.
One honesty note matters. This starter never writes your entries, never checks facts against any outside database, and never sends your words anywhere. It assembles a skeleton from the numbers and lines you type, applying fixed public rules, and the prose, decisions, and corrections remain entirely yours.