Player-Behavior-Driven Narrative Engine

You are the narrative systems director for [INSERT GAME TITLE]; design an AI story engine that continuously adapts quests and world reactions to observed player behavior. The system must deliver authored-quality emotional beats while generating variations consistent with lore, characters, and [INSERT NARRATIVE SCOPE]. Module 1 - Gameplay design: define narrative pillars, branch depth, agency model, story beats, action/story pacing, and endings policy. Module 2 - Generation pipeline: specify behavior-event ingestion, player-model inference, memory graph updates, beat planning, LLM drafting, lore/voice validation, and localization-aware templating. Enforce a safeguarded hierarchy: authored spine first, generated variations second, with hard gates on canon and character voice. Define session memory, long-term cross-session memory, forgetting curves, and player privacy/opt-out controls. Explain how conflicting co-op player actions are reconciled into one shared narrative state. Define authoring tooling that lets writers inject spine content and guardrails without code deploys. Specify evaluation using both synthetic player agents and human reader studies. Module 3 - Data structures: provide schemas for story beats, the memory graph, NPC relationships, player traits, generated quests, and consent flags. Module 4 - Testing: define coherence regression suites, voice-consistency scoring, contradiction detection, branch-reachability tests, and sensitive-topic safety eval. Module 5 - Operations: cover prompt/model versioning, review queues, telemetry, regional culturalization, kill switches, and rollback. Provide a text architecture diagram, sample event-to-quest traces, risk register, and KPIs for completion and resonance. List assumptions about the dialogue system, save model, and online/offline split. Begin with a one-page narrative design summary, then deliver modules one at a time for approval. ================