We built this because we needed PostgreSQL SQL parsing inside infrastructure tooling where CGO wasn’t viable (Alpine containers, Lambda, distroless images, etc).

The goal isn’t to replace Postgres server parsing it’s to give tooling structured query information (tables, joins, filters, CTEs, etc) without running a database or shipping native dependencies.

It’s been used internally for months and we recently cleaned it up enough to open source.

Happy to get feedback, especially around edge cases or weird SQL we might miss.