Inside dbt Wizard, the AI agent built to show its work
Blog post from dbt
dbt Wizard is a purpose-built AI agent for analytics engineering that aims to make AI-generated dbt changes easier to assess by showing their data and downstream impact, rather than only code diffs or chat histories. Presented at dbt Summit 2026, the tool treats dbt projects as connected systems, validates changes in development schemas before production, displays impacts through DAGs, surfaces assumptions, and uses independent validation processes to reduce errors; design partners reportedly saw 15–20% fewer production incidents. A live demo showed it handling competing definitions of “active player,” preserving parallel conversational contexts, detecting a potential metric bug, separating inexpensive code checks from data materialization, and creating pull requests with relevant context. Available through dbt Studio, a standalone platform workspace, and the command line, with a desktop app in private beta, Wizard also supports skills and MCP connections while isolating schemas across parallel work trees.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.