Characters: the cast list your dialogue builds, plus the sheets you write

In one line: this explains where the characters in a screenplay live now — the cast list builds itself out of the dialogue you type, the character sheets are .md files you write yourself, and the two find each other by name.

The character bible is gone; the job it did split in two

The old Script Studio had a Characters tab: a roster, an appearance matrix, five role types, four scope filters. None of that exists in a .script file. If you still have an old Script Studio work open, its Characters tab and its appearance matrix are exactly as they were — they just don't come across when you convert.

What it was actually doing was two different jobs, and 4.0 keeps them apart:

This is more than a move. In the old model a character was a record you created before you could use it (the app created it behind your back, which was its own kind of confusing). Now the name you type into the script is the character, and "a character who hasn't been created yet" is not a state that exists.

The cast list: whoever you typed is on it

Slima reads the character lines in every .script in the book and collects the names into one cast list for the whole book, ordered by dialogue lines, then by scene appearances. Change a line, add a name, and the list has already changed — there is no second copy to reconcile against your pages.

That list surfaces in four places:

1. The autocomplete on a character line. Anyone already in this scene sorts to the top, then a divider, then everybody else, each with its "N lines M scenes" count. When nothing matches exactly, the last row is Create new character "A-Tang" — and pressing it only types the name. No file is created. Both halves are ordered by dialogue lines, most first, and the counts are the whole book's, not this scene's.
2. The slash palette (type / on an empty line). Three groups: the five element types with their shortcuts — four in Taiwan format, which has no parenthetical — then quick-insert, which is this book's five most-used characters and five most-used locations, then New Scene and /note. Picking a location is what changes the scene heading. Same cast list behind the character half.
3. The Outline panel (Ctrl + Shift + O). With a .script open it has three modes: Scenes / Characters / Locations. Characters groups this episode's scenes by who is in them.
4. Cast → Character voices, in the toolbar once you switch to Preview — that button only exists in preview, which is where you would want a different reading voice per character anyway. Same list again.

Who is in this scene

The Board view of a .script (the Board button on the toolbar) is a column of draggable scene cards. Each card carries a row of character chips — that is the scene's cast:

• Solid chips were detected from this scene's own text. Hover one and it says Auto-detected from this scene's text (updates as you write).
• Dashed chips you added by hand: present but not named in the text — listening from the doorway, asleep, dead. The Character button adds one.
• Each chip has an ×. That is Remove from the list, and it does not touch your dialogue — it records "don't count this name in this scene". That is how you fix a false positive. Add the name back later and the exclusion lifts.

So the list is detected ∪ added − removed. The detected half always follows your text; the judgements you made by hand never go stale because you rewrote a line.

Character sheets: what is worth writing

The cast list knows a name, a line count and a scene count. If you want to write down what someone wants and what they lack, make a file: create a Characters/ folder, right-click it → New File, and name it after the character (Shen Yan.md). In the Ink Tide (Screenplay) sample the folder is already there: that book inherits the novel's whole reference shelf, so Characters/ arrives with eight files in it, one per character, next to Locations/ and Worldbuilding/. The novel, Ink Tide, has the same eight.

Once the filename matches the name in the script, that name in the script gets a dotted underline, and Ctrl (⌘ on macOS) + click opens the file. Locations work the same way: the location field of a scene heading links to a .md with that name. No matching file, no underline — you never get a dead link that does nothing.

As for what goes inside, this is the part of the old character bible most worth carrying over. Its fields were good screenwriting prompts:

Write this	Because
Want (the outward goal)	What they think they're chasing. The dialogue serves it.
Need (the inward gap)	What they actually lack. They usually can't say it, and the story makes them face it.
Skill / flaw	A protagonist's flaw is usually the root of the need, and the reason they fail.
Backstory	The past, the family, the one event that made them this. You don't have to finish it in one sitting.
Signature lines	The handful of lines that sound most like them. For keeping a voice consistent this beats any adjective.

Characters/Shen Yan.md in the Ink Tide novel sample is exactly this shape: Want is "to find out what became of her father after the exile", Need is "to admit she is a word-waker, and stop apologizing for her father's crime" — with appearance, temperament, flaw, ability and a twelve-chapter arc under them.

Renaming and deleting behave differently now

Renaming is not global. In the old bible, renaming a character rewrote the dialogue with it, wiki-link style. Now the name is the identity, so editing one line edits one line. To change a name across the book use search and replace.

Deleting no longer cascades. Deleting an old character was a hard delete that unhooked every reference in the script. A character sheet is just a file — put it in the trash (trash and restore) and not one word of the screenplay changes. There is exactly one way to get someone off the cast list: take their lines out of the script.

Character templates no longer have a way in. They were saved sets of custom fields you could stamp onto another character, and they still work inside an existing .character file — but you can't create one of those any more, so in practice nobody meets them. A sheet is a .md now, so to reuse a structure you duplicate a file you already like.

The appearance matrix: no honest replacement

The old matrix was a character × episode grid — one look told you the protagonist vanished in episode 4, or that the antagonist doesn't appear until episode 3. A .script file has no such grid — it lives in the old Script Studio, and an old work still shows it. What a .script gives you are pieces of it:

• The card behind the magnifier beside a character line (View character info): it reads N scenes · N lines, with a per-episode distribution in bars underneath — the closest thing to one row of the matrix.
• The book-wide cast list ordered by dialogue lines, which tells you who carries the show and who has two lines.
• Outline → Characters, which groups characters against scenes inside one episode. It does not cross episodes.

If you genuinely need the grid, you have to keep it yourself for now — a file card per episode on a .beats board, or a table in a planning .md. That is a step backwards, and it is written here so you don't spend an afternoon hunting for a tab that isn't there.

Related

• Docs: Dialogue and character autocomplete
• Docs: Scene headings: INT / EXT and locations
• Docs: What you see when you open a screenplay file
• Docs: How AI helps with a screenplay
• Docs: File tree overview
• Docs: Manuscript vs reference
• Docs: Relationship maps
• Docs: The old Script Studio: what still works