Story Structure
Story Bible
Check for internal contradictions in character attributes and scene lighting/time of day.
What It Does
Extracts and cross-references two types of narrative facts:
- Character attributes — physical states, limitations, and abilities mentioned for named characters
- Scene lighting/time of day — references to whether it's light or dark, morning or night
Flags contradictions within or between scenes.
Why It Matters
Continuity errors break the reader's trust. If a character "broke her right arm" in Chapter 3 and "swung the bat right-handed" in Chapter 5, careful readers will notice. If a scene starts "under the midday sun" and two paragraphs later someone "squinted through the darkness," the reader loses spatial orientation. The Story Bible analyzer automates the eagle-eyed beta reader who catches these inconsistencies.
What Gets Flagged
Character Attribute Contradictions
Severity: Warning
Example (flagged):
Story Bible: character "Elena" is described as having a physical limitation ("broken wrist") but later performs a vigorous action ("climbed the rope hand over hand")
Why: The character's stated physical state contradicts their subsequent actions.
Scene Lighting Contradictions
Severity: Information
Example (flagged):
Story Bible: scene lighting conflict — "afternoon sunlight" established but "pitch darkness" mentioned later in the same scene
Why: The time of day or lighting conditions change without explanation within a single scene.
What Gets Tracked
| Category | Examples of tracked attributes |
|---|---|
| Physical limitations | broken, injured, blind, deaf, limped, wheelchair |
| Physical actions | ran, climbed, fought, swung, carried |
| Light indicators | sunlight, moonlight, dawn, dusk, darkness, lamplight |
| Time markers | morning, afternoon, evening, midnight, noon |
Configuration
No configuration options.
Technical Details
- Source:
prose-craft - Scope: Document-level (cross-references attributes across scenes)
- Method: Named entity extraction via capitalized word detection; attribute assignment via regex; contradiction detection via rule-based conflict checks