Rules

Rules add business context to raw column definitions. They enforce ranges, formats, and string/object constraints inline with each field.

Column rules via the CLI

Required/optional: dbx schema field <aggregate> <field> --required or --not-required.
Formats: --format email|url|credit_card|country_code|iso_8601|wgs_84|camel_case|snake_case|kebab_case|pascal_case|upper_case_snake_case.
Contains / does_not_contain: --contains foo / --does-not-contain bar.
Regex: --regex '^[A-Z]{2}-\\d+$' (repeatable).
Length: --length-min <n> --length-max <n> for text/binary.
Range: --range-min <value> --range-max <value> for numeric/date/time.
Object properties: --properties '{"address":{"type":"text","required":true}}' to nest rules under an object field.

Use fixtures before promoting to strict mode:

dbx schema validate person --event person_registered --payload @fixtures/person.json

Combine with CI to block changes that break validations.

In restrict=default, events without schemas are allowed; declared schemas validate and return clear errors when violated.
In restrict=strict, every aggregate must have a schema and any rule violation rejects the write.

Use this feedback loop to educate producers about expectations while giving you the option to enforce them once confident.