Skip to content

CLI Reference

playground

Create a self-contained project that loads CSV files into DuckDB, and change its data for an incremental load.

sqlload playground [NAME] [--template shop]
sqlload playground --change

sqlload playground writes a project into a new folder, shop by default. It loads 50 customers and 1,000 orders from CSV files in the project’s data/ folder into the DuckDB file shop.duckdb, so it needs no database, no Docker and no credentials. The Quickstart runs it from start to end, with its output.

File What it is
sqlload.toml The project: a filesystem connection to data/, a DuckDB target, and one files source per folder, each merging on its key.
data/customers/customers.csv 50 customers.
data/orders/orders_2026-11-01.csv 1,000 orders.
.gitignore Leaves shop.duckdb out of Git.

The folder must not exist yet or must be empty; otherwise the command fails and changes nothing.

sqlload playground --change, run in the project’s folder, adds the next file of changed orders to data/orders/ and prints each order it changes or adds. The next sqlload load merges only that file. With --config, it changes the playground whose sqlload.toml that is.

Once every change is in the project, --change adds nothing and says so. To start again, delete shop.duckdb and the added files; the next sqlload load loads everything again.

Option Meaning
NAME The new project’s folder. Default shop.
--template <TEMPLATE> The project to create: shop (default).
--change Add the next change to an existing playground’s data instead of creating one. Takes no NAME or --template.