Check a dataset for missing values, duplicated rows, possible numerical outliers and geometry-type consistency.
Review flags against source definitions and your intended use. These diagnostics cannot establish that a dataset is complete, representative or fit for purpose.
Choose whether you need to inspect a dataset, check for source updates or export package data. These are independent utilities, not an automatic maintenance pipeline. Review detected issues or updates, record provenance and versions, and include definitions and reuse information when sharing data.