Building Dbt Semantic Layer

This one handles the full lifecycle of dbt's Semantic Layer: semantic models, entities, dimensions, and all five metric types (simple, derived, cumulative, ratio, conversion). It's smart about routing you to the right YAML spec based on your dbt version, and if you're on 1.12+ it'll offer to migrate legacy configs with dbt-autofix. The guide works backwards from business questions or forwards from existing models, which makes it flexible whether you're starting fresh or retrofitting analytics onto established tables. One caveat: cumulative metrics need a time spine configured separately, and the spec has some gotchas around grain-to-date versus window parameters. Solid reference if you're moving beyond raw SQL aggregations into governed metric definitions.

Install to Claude Code

npx -y skills add dbt-labs/dbt-agent-skills --skill building-dbt-semantic-layer --agent claude-code