Using dbt for Analytics Engineering

What it does

This skill equips an agent with best-practice guidance for working with dbt: planning and building models, using ref() and source(), writing tests, debugging errors, and validating results via dbt show. It emphasizes software-engineering principles (DRY, modularity, testing) applied to data transformations and includes reference guides for planning models, discovering data, writing tests, and debugging.

When to use it

Use this skill when constructing new dbt models, refactoring or extending existing models, creating analytics pipelines, writing data tests, or investigating failures in dbt runs. Avoid using it for semantic-layer question answering; use dedicated skills for that.

What's included

• Scripts: yes (project scripts/ present)
• References: detailed reference docs under references/ (planning, discovering data, tests, debugging)
• Instructions: step-by-step guidelines for model planning, using {{ ref }} and {{ source }}, running dbt show for validation, cost-management tips, and CLI interaction patterns.

Compatible agents

Likely compatible with agents that can run shell commands and interact with CLIs (Copilot/Codex-like agents, Kilo CLI-enabled agents).