MotherDuck integration for Rig
The easiest way to query MotherDuck in Claude and other AI tools
MotherDuck is a cloud data warehouse built on DuckDB that runs fast analytical SQL on data in the cloud and on the laptop.
What Rig sees in MotherDuck
Once connected, Rig models these objects so questions resolve to the right tables and joins:
- Databases
- Schemas and tables
- Views
- dbt models
- Parquet and CSV attachments
- Query history
How MotherDuck connects to Rig
Rig connects to MotherDuck (and local DuckDB) with a read token, reads the schema, and builds a context layer over your tables, joins and metrics. Queries run as DuckDB SQL, which makes it well suited to fast, exploratory questions.
Run fast exploratory analytics in plain English
DuckDB's speed makes MotherDuck a strong fit for back-and-forth questioning, and Rig turns each plain-English question into validated DuckDB SQL. You can iterate on a cohort or funnel question in seconds without writing the SQL by hand.
- What is the seven-day retention by signup cohort?
- Show revenue by week from the MotherDuck warehouse
- Which channels drive the most activated users?
Query Parquet and CSV files alongside warehouse tables
DuckDB reads Parquet and CSV directly, and Rig treats those attachments as part of the context layer. You can join a file dropped into MotherDuck against your modelled tables without a separate load step.
- Join the uploaded campaign CSV to orders
- Summarise the Parquet export by month
- Compare the file totals against the warehouse table
Serve MotherDuck to Claude, Cursor and ChatGPT
Rig exposes MotherDuck to AI assistants over MCP with your governed schema attached, so the assistant answers from agreed definitions. This pairs well with analysts who prototype in DuckDB and want the same numbers in chat.
- Pull active users by plan from MotherDuck
- Find the join path from sessions to conversions
- Explain how the activation metric is built
Keep DuckDB queries validated and bounded
Rig validates every query in a sandbox before it runs, so even fast iterative questioning stays safe. Read-only access keeps the warehouse protected while analysts explore freely.
- Which queries returned the most rows this week?
- Add a limit and re-run that query
- List the tables available in this database
Ask your MotherDuck data in Claude, ChatGPT or Cursor
Rig serves your MotherDuck data to AI assistants over MCP. Once MotherDuck is synced into your warehouse, you can ask questions in plain English from the tool you already use, and every answer is backed by validated, governed SQL.
Frequently asked questions
- Does Rig work with local DuckDB as well as MotherDuck?
- Yes. Rig speaks DuckDB SQL, so both MotherDuck in the cloud and local DuckDB databases connect to the context layer.
- Can Rig query Parquet and CSV files in MotherDuck?
- Yes. DuckDB reads Parquet and CSV directly, and Rig can join those attachments against your modelled tables.
- Can I ask MotherDuck questions in Claude or ChatGPT?
- Yes. Rig serves MotherDuck to AI assistants over MCP for plain-English querying.
- Is access to MotherDuck read-only?
- Yes. Rig connects with a read token and sandbox-validates every query before it runs.
Related integrations
Connect MotherDuck data to AI tools like Claude
Rig builds a governed context layer over your data so every team, and every AI tool, asks questions and gets answers they can trust.