Use the Test Data Generator to make a small batch of plausible records, then copy the chosen records into a checked test fixture. Each click produces new random values, so the tool is useful for preparing data and exploring layouts, not for reproducing the same values on every CI run.
\n\nA form test fails only when a long name wraps onto a second line. Re-running it with newly generated data changes the input again, and the failure disappears. The data is useful because it revealed a boundary; the test becomes useful only after the boundary value is saved.
\n\nOpen the tool, select Full Name, UUID, and Job Title, set the row count to two, and press Generate. Switch between JSON and CSV to choose the format your fixture loader accepts. The generator produces two records with those three keys. On the live page, one run produced:
[\n {\"name\":\"Sofia Agarwal\",\"uuid\":\"344d7e85-5527-4382-80c3-2ea219dbacef\",\"job_title\":\"QA Engineer\"},\n {\"name\":\"Ava Williams\",\"uuid\":\"4919f380-de38-4786-8b49-f8a4331db841\",\"job_title\":\"Backend Engineer\"}\n]Your names and IDs will differ. The CSV export has the same columns: name,uuid,job_title. The page can copy or download the current output; switching format does not generate a new batch.
The implementation uses Math.random() for names, IDs, and most other fields. It has no seed control. If a generated record exposes a bug, copy that exact record into a fixture and give the test an assertion about the observed behavior. Do not ask the tool to regenerate it next run.
The field list includes email, phone, password, and card-shaped values. Treat them as display and validation samples only. The generated email domain can be a real mail provider, and the password generator is also based on Math.random(); neither belongs in a live account or a security test. Use reserved addresses and a dedicated test payment provider when the workflow could send mail or process a card.
Two plausible people will not reliably produce an empty string, a maximum-length name, or a duplicate ID. Add those cases by hand. Use the generator for ordinary records and use deliberate fixtures for the conditions that make a test meaningful.
\n\nIs the output repeatable? No seed is exposed. Save the exact rows you want to run again.
\n\nCan I download 1,000 rows? The page's row-count control tops out at 100. For larger, repeatable datasets, write a seeded generator in the test project and keep its seed with the run evidence.
\n\nGenerate a handful of records for one form. Save one ordinary case and one manually chosen boundary case as fixtures. Record the failing input with the test result so the next run exercises the same condition.
\n\n