Day-Zero Data Quality Gatekeeper — manifest-driven pre-ingestion checks built on GX

[huacenxu]

We built an open-source tool called Day-Zero Gatekeeper on top of GX 0.18.x for teams who need data quality checks before ingestion — with zero historical data required.

The approach: encode the full data contract in a YAML manifest (schema, health, business rules) and gate the pipeline on Day One, before any history exists.

We added a Layer 4 completeness check that GX doesn't have natively — compares source-to-raw record counts via a vendor control file, with a same-vendor taxonomy proxy fallback when no manifest is available.

Sharing as a community extension for anyone building pre-ingestion DQ pipelines.

👉 https://github.com/huacenxu/day-zero-gatekeeper

[joshua-stauffer]

Thanks for sharing @huacenxu.

One important caveat for anyone considering adopting it: this tool is built on GX 0.18.x, which is a legacy version, and no longer receives updates or bug fixes. If you're starting a new project, we'd strongly recommend building against the current GX release (docs here) to get active support, security patches, and the current API.

@huacenxu looks like the lift would be low to migrate to the v1 API, but let me know if you have questions.