JSON Schema
Waiting for input.
Mock data
Respects enum, const, formats, numeric and length bounds, required vs optional, and $ref to local pointers.
Respects enum, const, formats, numeric and length bounds, required vs optional, and $ref to local pointers.
Tests and demos need data that looks real without being real. Feed this a JSON Schema and it produces records that satisfy it — respecting enums, const, numeric and length bounds, and formats such as email, uuid and date-time. Field names are used as hints too, so a field called "city" gets a city rather than a random word. The seed makes it repeatable: the same schema and seed always produce the same data, which is what makes the output usable in a committed fixture.
Feed it a JSON Schema and get records that satisfy it, ready for tests, demos and stubs.
The same schema and seed always produce the same data, so output can live in a committed fixture.
A field called email looks like an email, city like a city, createdAt like a timestamp.
Enums, const, numeric and length bounds, array sizes, required versus optional, and local $ref.
By design — the generator is seeded rather than random. Change the seed for a different set; keep it to get a byte-identical result you can commit as a test fixture.
It respects enum, const, formats, minimum/maximum, minLength/maxLength, minItems/maxItems, uniqueItems, required versus optional, and $ref to local pointers. Regex patterns are not solved — a pattern-constrained string gets a placeholder.
Up to a hundred records per run in the picker. For more, generate and repeat with different seeds, or convert the result to JSONL for line-based tooling.