entity-tracker

Track manuscript entity continuity by maintaining a persistent structured ENTITY_STATE.yaml that extracts and tracks characters, locations, objects, and world rules. It serves as a single source of truth to eliminate continuity drift across long-form book production.

Extract and update narrative entities by processing specified manuscript chapters incrementally in UPDATE mode. This extends or revises tracked entities in the YAML state file without rebuilding the entire entity history from scratch.

Yes, you can manage worldbuilding rules and timeline entries using a YAML schema. The process requires strict YAML schema conformance to record narrative entities, timeline entries, objects, and world rules extracted from manuscript chapters.

Detect continuity conflicts and knowledge gaps by flagging contradictory facts during chapter extraction. The system records these as unresolved conflicts or knowledge gaps for downstream continuity review without automatically fixing them.

Build a persistent source of truth by running the entity-tracker in BUILD mode. It generates the ENTITY_STATE.yaml file by extracting and merging narrative data from your existing foundation.md, outline.md, and all current chapter files.

Yes, entity tracking works for incremental chapter revisions during long-form book production. The system applies chapter-scoped extraction with deterministic sources and incremental merge logic to update the YAML state without processing the entire manuscript.