writer-memory

Persistent writer memory tracks characters, relationships, scenes, and themes across sessions using JSON-backed local storage. It requires initialization to establish a project memory directory, then uses deterministic CRUD commands to store and retrieve character emotional arcs, tones, and speech levels.

Dialogue voice validation checks new dialogue against established character constraints, including tone, speech level, keywords, and taboos. It synthesizes the stored memory state to flag accidental inconsistencies, helping you verify that character voices remain coherent throughout the story.

Emotion-first synopsis generation creates a synopsis structured around emotional coherence, mapping attitude to relationship structure, emotional theme, genre versus real emotion, and ending aftertaste. It verifies narrative coherence while drafting by ensuring the emotional arc aligns with established scene tags.

You can track evolving relationships and scene continuity by storing relationship types, dynamics, and events alongside scene cuts, narration tone, and emotional tags in a local writer-memory directory. This persistent context captures world rules and atmosphere to maintain continuity.

Korean emotion vocabulary tagging is integrated into the fiction context tracking system, applying specific emotional tags to scenes and characters. This allows Korean storytelling workflows to maintain emotional continuity and validate dialogue using established Korean speech levels.

No external dependencies are required to set up persistent memory for creative writing. The system operates independently using a local JSON-backed storage directory, requiring only project initialization to begin adding characters, relationships, scenes, and themes through deterministic commands.