fivetran/great_expectations — reverse-engineered prompt

Build me a Python library for data quality checks, something like Great Expectations. I want data teams to be able to describe what good data should look like in plain, reusable rules, then run those checks against datasets and see clear validation results.

The core flow should feel simple. I should be able to install it, import it in Python, create a project or data context, define expectations for things like required columns, value ranges, null limits, uniqueness, and basic schema rules, then validate data and get readable output. Please also generate human friendly documentation for validation results so non engineers can review what passed and failed.

Keep it centered on Python and modern versions people actually use. Make the package feel production ready, with tests, docs, examples, and a clean developer experience. If there are integration details or compatibility questions, look up the current docs online if you need to. I want something open source friendly and easy for data engineers and analysts to adopt.