Editing Data

Edit right in the result

Double-click any cell in a Result node to edit it in place. Change the value, press S to commit, or esc to back out. Peek builds the UPDATEfor you, runs it, and re-runs the query so the cell you're looking at always reflects what's actually in the database.

Editing is keyed off the primary key, so it works on any single-table SELECT from a table with a primary key. If a query can't be safely edited - a join, an aggregate, a table with no key - Peek tells you inline instead of writing something ambiguous.

Schema-aware inputs

Peek understands your column types and gives you the right control for each one. A boolean becomes a TRUE / FALSE / NULL dropdown, a JSON or JSONBcolumn opens a multi-line editor, and numeric columns get a number input. There's a dedicated NULL button to clear a value, and types are cast correctly as the statement is built.

You can even reference a variable while editing - type @name in a cell and Peek substitutes the value from any connected Variable node before writing.

Insert and delete, deliberately

Add new rows straight from the Result node, with the same schema-aware inputs you get when editing. Select one or more rows and delete them too - but only after a confirmation step that shows exactly which table and how many rows are about to go, so a destructive write is never a single misclick.