dbt validator

The tool provides deterministic validation for dbt projects by executing core dbt commands to verify model syntax, compilation, and dependency resolution. It is designed for data engineers, analytics engineers, and CI teams who need to prove that a project can be built and tested consistently before deployment, without performing data loads. It reports configuration issues, missing profiles, or incompatible adapters early.
At runtime, the validator loads the project and profiles, resolves dependencies, and runs targeted commands (parse/compile, deps, test, and optional run) to generate a structured result. It collects counts for models, tests, and failures, along with failure traces. The output is designed for automation and integration into CI pipelines, pull requests, and pre-merge checks.
Unique value lies in producing a reproducible, machine-readable report from a validation path that does not require a data warehouse connection or data processing. Users benefit from early feedback on schema changes, model references, and test expectations, reducing debugging time and ensuring consistent releases across environments.
Core features include: project discovery, profile handling, targeted model selection, sequential or parallel execution, and a JSON-ready summary with detailed failures. Typical use cases include PR validation, nightly quality checks, and onboarding of new dbt users. The tool emphasizes transparency by exposing fail lists, logs, and execution durations.