What problem does it solve?
Helps data engineers and analytics engineers create, migrate, and maintain dbt Semantic Layer artifacts so metrics, entities, dimensions, and time-based analyses are correct and discoverable across projects.
Core Features & Use Cases
- Spec-aware authoring: Detects and guides whether to use the latest model-annotated spec or the legacy top-level semantic_models spec and recommends upgrades when appropriate.
- Metric and model design: Walks through semantic_model or semantic_models creation, primary entity selection, dimension discovery (including time dimensions), and simple/derived/cumulative/ratio/conversion metric authoring.
- Validation and tooling: Requires running dbt parse or dbtf parse and then semantic validations (dbt sl validate or mf validate-configs), enforces time-spine requirements, and prevents common pitfalls like filtering on non-declared columns.
- Use case: Add a semantic layer to an orders model, define the primary entity and time dimension, create order_count and total_revenue metrics, and validate with MetricFlow.
Quick Start
Enable or migrate the semantic layer for a model by adding semantic_model metadata, define the primary entity and time dimension, add simple metrics, then run dbt parse and mf validate-configs to validate.