Travailler sur le projet dbt ZLV

To build and test dbt models with DuckDB and MotherDuck, this Skill provides step-by-step guidance for creating, modifying, and validating models across staging, intermediate, and marts layers. It enforces reproducible steps and guardrails for safe, auditable dbt development in team settings.

Structuring a dbt project for team collaboration involves using reusable playbooks and contextual references to maintain consistency across staging, intermediate, and marts layers. This Skill automates guidance for AI agents, providing templates for adding models, writing tests, and validating dependencies.

Yes, you can use AI agents to automate dbt model development and documentation. This Skill applies AI-guided workflows to explore, modify, and validate models, ensuring reproducible steps and safe, auditable development within the zlv_dbt_project.

To add a new marts model and validate its dependencies in dbt, you provide the AI with a concise, goal-oriented action. The Skill supplies templates for adding models, writing tests, and validating with dbt docs to ensure reproducibility and correct dependency resolution.

This dbt workflow supports reproducible and auditable model testing by enforcing guardrails and contextual references for safe development. It provides quick access to project structure, macros, and tests to validate models across stg, int, and marts layers using DuckDB and MotherDuck.

You should not use an AI-guided workflow for dbt development if your project lacks clear guardrails or relies on unsupported platforms outside of DuckDB and MotherDuck. The Skill enforces reproducible steps specifically tailored for the zlv_dbt_project architecture and team collaboration.