SCRIPTA Documentation Index

Narrative Theory & Story Elements

Comprehensive documentation on characters, themes, plot elements, narrative structures, world rules, wisdom patterns, and quality metrics.

Characters Themes Plot Narrative World Rules Wisdom Patterns Metrics

DS01 Vision and Strategic Goals

Problem context, 5 design principles, strategic goals with justification and limitations, KPIs, and risk analysis.

Raw .md

DS02 System Architecture and API

Pipeline architecture, data model, REST endpoints, authentication mechanisms, and repository structure.

Raw .md

DS03 Research Framework and Evaluation

Research questions, 6 experimental variants, metrics definitions, and validation protocol.

Raw .md

DS04 CNL Specification

SVO CNL purpose, lexical rules, statement and group syntax, entity types, constraints, and pattern system.

Raw .md

DS05 VSA/HDC Specification

Hyperdimensional computing theory, parameters (10K dimensions, bipolar), operations (bind/bundle/permute), and algorithms.

Raw .md

DS07 Use Case Definition

Problem scope, 5 actors, primary workflow, alternative workflows, and traceability requirements.

Raw .md

DS09 Story Studio Interface Specification

UI layout, palette, canvas interactions, context panel, generation/evaluation panels, and data flow.

Raw .md

DS10 Visual Story Composer Specification

Visual tree editor paradigm, libraries, pattern instantiation, and rules for generating CNL from structure.

Raw .md

DS11 CNL Syntax Reference

Complete SVO CNL syntax reference and unified AST structure.

Raw .md

DS12 Metrics Interpreter and Evaluation Suite

Execution model for SVO CNL, metric plugin runtime, human evaluation rubric, and statistical analysis protocol.

Raw .md

DS13 Metric — Coherence Score (CS)

Formal definition and deterministic computation for CS using entity continuity, causal heuristics, and penalties.

Raw .md

DS14 Metric — Character Attribute Drift (CAD)

Trait drift measurement via embeddings over token windows with per-character breakdown and thresholds.

Raw .md

DS15 Metric — Compliance Adherence Rate (CAR)

Guardrail pass rate definitions (strict/lenient), artifact status model, and corpus-level reporting.

Raw .md

DS16 Metric — Originality Index (OI)

Distance from a curated trope corpus using deterministic embeddings, including top-k nearest tropes.

Raw .md

DS17 Metric — Emotional Arc Profile (EAP)

Correlation between measured emotional trajectory (moods/text) and target arc templates.

Raw .md

DS18 Metric — Retrieval Quality (RQ)

Semantic retrieval evaluation using Recall@k and MRR on held-out queries with relevance labels.

Raw .md

DS20 Metric — CNL Parse Success Rate (CPSR)

Translation success rate from NL to valid SVO CNL with parser + semantic validation criteria.

Raw .md

DS21 Metric — Constraint Satisfaction Accuracy (CSA)

Scope-aware evaluation of SVO constraints (requires/forbids/must/tone/max/min) with evidence.

Raw .md

DS22 Metric — Explainability Score (XAI)

Human survey instrument (Likert 1–5), aggregation, confidence intervals, and acceptance thresholds.

Raw .md

DS23 Metric — Narrative Quality Score (NQS)

Composite score combining CS with human ratings, plus +25% improvement criterion vs baseline variant.

Raw .md

DS24 Architecture Overview

Browser-first architecture, persistence-only server, module structure, and SDK vs Node/CLI boundaries.

Raw .md

Evals Dataset

NL to CNL evaluation pairs for SCRIPTA testing and benchmarking.

Raw .md