Getting seed rows as INSERT statements instead of typing them
Filling a table with enough rows to render a screen usually means copying the same INSERT statement dozens of times and changing one number each round. Enter the table name, the column definitions and how many rows you want, and runnable INSERT statements appear on the spot. Write one column:type pair per line.
Every value is deterministic. int counts from 1, decimal steps by 1.5, date starts at 2026-01-01 and advances a day per row, datetime advances an hour, and bool alternates TRUE and FALSE. String values combine the prefix you type with the row number, and single quotes are doubled as the SQL standard requires. Table and column names are validated against a whitelist of letters, digits and underscore; anything else is rejected instead of being wrapped in quotes. Backslashes are blocked at input because databases read them differently.
The limits matter. Foreign key references, constraints beyond NOT NULL, auto-increment collisions and dialect-specific functions are not handled. The output is seed data for development and does not belong in production as-is. Written as of October 2026.
Frequently asked questions
Single quotes are doubled ('') as the SQL standard requires. Backslashes are rejected at input because databases read them differently.
No. Table and column names are checked against a whitelist of letters, digits and underscore. Anything else is rejected rather than wrapped in quotes.
Not advisable. The values are sequential seed data for development, so review your constraints and target schema before running anything.