Bring the draft.Keep everything you wrote.

What you can bring

Drop in single files, several at once, a whole folder, or a zipped folder, and Authrot reads all the way down. Folder names help: a folder called Characters or Story Bible sends its files to the bible, and a folder named like Book Two starts a new book. Clutter such as system files and hidden folders is skipped quietly.

A few formats need a step first. Google Docs come in as a Word download or a paste, which keeps italics and bold. Apple Pages, old .doc and .odt files should be exported to Word. A PDF needs a real text layer, so a scanned book is refused, and a PDF doesn’t bring italics back. EPUBs locked with DRM can’t be opened.

• Word (.docx), and Google Docs downloaded as Word or pasted
• Markdown and Obsidian-style vaults, front matter included
• Rich Text, PDF, EPUB, HTML and plain text
• Zipped or dropped folders, read in order

Read on your machine, not ours

Your files are parsed in your browser, never on Authrot’s servers. What reaches the studio is the text you choose to import, after you’ve checked how it was split. For a writer bringing in an unpublished manuscript, that matters: the file itself never leaves your computer.

Before anything else, the importer asks what the files are: your manuscript, your notes and ideas, or both. It sounds like a small question, but it prevents the most expensive mistake an importer can make. A page of brainstorming filed as a chapter would be read as if it had happened, maybes and all. Tell it they’re notes, and they stay notes.

You can import into a project you already have, from the manuscript section of the sidebar, or tick “Start from an existing manuscript” when you create a new one. In a new project, the empty first chapter it starts with is replaced by your own, so you don’t end up with a blank chapter one in front of your real one.

Review the split before anything is saved

A manuscript file rarely says plainly where its chapters begin. The importer looks for chapter markers in every common style (Chapter 7, CHAPTER SEVEN, Chapter XII: Salt, Prologue, Epilogue), recurring headings, and page breaks, and proposes sections. Each one has a title, a destination (chapter, story bible, outline, note, brainstorm or skip), a part kind such as prologue or interlude, and a word count.

You can join a section to the one above, split one at any paragraph, or send a whole file to one destination. Title pages and table-of-contents lines are proposed as skip. Reading order lists every chapter across every file, to drag into place before you import. Nothing is written until you press Import.

• Front matter read for titles, order, book and summary
• Chapters placed by hand stay pinned where you put them
• ✦ Let AI sort this, from each section’s opening, never the whole text

Why reading order matters

Chapters are read into the bible in order, each one against the ones before it. That’s how Authrot knows what has happened by chapter nine, which characters have gone quiet, and which setups are still waiting to pay off. A chapter imported in the wrong place gives every chapter after it the wrong story so far.

So the importer is careful about order. Numbers in titles and file names sort naturally (chapter 2 before chapter 10), front matter can set an order, and when there’s nothing to go on, it tells you to check. If you use the AI, Let AI sort this can propose an order from each chapter’s opening and a passage from its middle, around the chapters you’ve already placed, and Undo puts everything back.

From a pile of files to a working studio

Two options decide how much happens after import: building story-bible entries with the AI, and indexing chapters. With both on, character sheets and world notes become entries with their aliases and traits, wiki-style links in Markdown become hints for which entries exist, and each chapter is read so the bible, threads and promises reflect the story you’ve already written. You see an estimate before you start.

With them off, nothing reads your text: chapters, notes and outlines simply land in the project, and you can build the bible by hand, link names yourself, and have chapters read later. Either way, you can start writing straight away.

If you’re unsure, a sensible middle path is to import everything with the AI off, check that the chapters and their order look right in the manuscript, and then have the chapters read, first to last, once you’re happy. Reading in order from a clean import gives the bible the right story so far for every chapter.

Importing again, safely

Writers often keep editing the original file for a while, or get a marked-up copy back from an editor. Drop the newer file in, and Authrot matches each section to the document it made last time and tells you whether it’s unchanged, changed in the file, or edited here since. You choose per section: replace, keep yours, import a separate copy, or skip.

When a chapter is replaced, the version from before is saved first, so your text is always a restore away, and the chapter keeps its place, its links and its history. Nothing in the project is removed because it isn’t in the new file. Story-bible files that haven’t changed aren’t read again, so a re-import doesn’t pay twice for the same notes. If you’d like to review a whole re-import before it touches the book, make a branch, import there, and merge when you’re satisfied.