What problem does it solve? Designing analytical data models in MotherDuck often results in ad-hoc SQL executed directly in the warehouse with no versioning, no dependency tracking, and no reproducibility. This Skill produces a file-based project scaffold with organized SQL models and a DAG manifest so transformations can be reviewed, versioned, and rerun. ## Core Features & Use Cases - File-Based Project Scaffold: Generates SQL files organized by lifecycle stage (raw, staging, analytics) plus a model_manifest.yml defining the DAG, dependencies, and materialization strategy. - Analytical Schema Design Guidance: Applies OLAP best practices such as wide denormalized tables, pre-aggregated serving tables, correct data types (DECIMAL for money, VARCHAR for IDs), and NOT NULL constraints. - Use Case: When asked to model raw event data for analytics, the Skill inspects source tables, then creates raw/staging/analytics SQL files with a manifest declaring each model's dependencies and materialization, ready to run against MotherDuck. ## Quick Start Ask the assistant to design a data model for your MotherDuck source tables and generate the SQL project scaffold with a dependency manifest.
