MotherDuck

MotherDuck is hosted DuckDB. Use it when the destination should be shareable without running DuckDB on the Skipprd host.

Skippr, your data department

Speak to an expert. Upfront insights in days, not months.

Book an intro call →

20 minutes · No sales pitch

Before you begin

• A MotherDuck token (motherduck.com → settings).
• A database name. Skipprd creates table in schema when it can.

export MOTHERDUCK_TOKEN="..."

Configure

Python

from skippr import Config, DataSinkMotherduck, EnvRef, Pipeline

cfg = Config.discover()
md = cfg.data_sink(
    "warehouse",
    DataSinkMotherduck(
        motherduck_token=EnvRef("MOTHERDUCK_TOKEN"),
        database="analytics",
        schema="bronze",
        table="events",
    ),
)
cfg.pipeline("events", Pipeline(data_source=cfg.get_data_source("sample"), data_sink=md))
cfg.save()

CLI

skipprd connect data-sink motherduck \
  --pipeline events \
  --name warehouse \
  --motherduck-token '${MOTHERDUCK_TOKEN}' \
  --database analytics \
  --schema bronze \
  --table events

YAML

data_sinks:
  warehouse:
    Motherduck:
      motherduck_token: ${MOTHERDUCK_TOKEN}
      database: analytics
      schema: bronze
      table: events

Options

Key	Type	Required/Default	Description
motherduck_token	secret	Required	Token as ${ENV}
database	string	Required	Database
table	string	Required	Table
schema	string	Not set	Schema (MotherDuck default if unset)

How data lands

Retries are exactly once. For a local file warehouse, use DuckDB instead.

Troubleshooting

Symptom	Fix
Auth failed	Recreate the token and export it in this shell
Database missing	Create it in MotherDuck, or use a token that can create databases

Next steps

• MotherDuck source
• DuckDB