Story Index
A structured cast list with arcs, voices, and relationships. Different from the World Index in that entries are characters and story-specific entities (artifacts, organizations tied to the plot) rather than universe-wide standing facts.
Scope levels
Each entry lives at one of three scopes:
- Universe, visible to every project in the workspace. Useful for shared-universe characters.
- Series, visible to every project in a series. The default for series-spanning protagonists.
- Project, visible to one project only. One-book characters.
Entry fields
- Name + aliases, primary identifier, plus aliases the highlighter recognizes in prose.
- Type, character, faction, object, location, concept, etc. Drives the entry's color in the editor highlighter.
- Description, the canonical fact set.
- Voice seed, speech patterns, hedge words, vocabulary range. Used by the AI when generating dialogue for this character.
- Transformation arc, how the entry changes across the book.
- Promise hooks, connections to the project's Promise (what the entry is set up to pay off).
- Importance, 0-100 weight that influences which entries the AI's context bundle prioritizes.
In-editor highlighting
Once an entry exists, the prose editor highlights every occurrence of its name and aliases, colored by the entry's type. Click a highlight to open the entry. The AI context bundle uses these highlights to figure out which entries are present in the current scene; it won't invent attributes for entries it can see in the text.
Best practices
- Add an entry the first time a character is named. It's cheap and prevents the AI from inventing biographical detail later.
- Set voice seeds early. A line in the seed (“Hedges with ‘I think’ before any strong opinion”) steers AI dialogue more than a paragraph of personality description.
- Keep aliases tight. A character with 12 aliases will saturate the highlighter and hurt context bundle quality.