Skip to content
| Marketplace
Sign in
Visual Studio Code>Data Science>DaffyDNANew to Visual Studio Code? Get it now.
DaffyDNA

DaffyDNA

trobeeone

|
3 installs
| (0) | Free
A SQL Server-compatible warehouse in VS Code: query folders of parquet, CSV, JSON and Excel files, and SQL Server, PostgreSQL and MySQL databases, with T-SQL from mssql, SSMS or Power BI. Runs DaffyDB, on DuckDB.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

DaffyDNA

A SQL Server-compatible warehouse in VS Code. Query folders of Parquet, CSV, JSON and Excel files, and SQL Server, PostgreSQL, MySQL, SQLite and DuckDB databases, with T-SQL: from Microsoft's SQL Server extension (mssql), SSMS, Power BI, or your own code. One query can join last month's Parquet export to a table in SQL Server and another in PostgreSQL.

DaffyDNA runs DaffyDB, which runs on DuckDB. DaffyDB is like Babelfish for PostgreSQL, which lets SQL Server clients work with PostgreSQL, but for DuckDB: it speaks SQL Server's wire protocol, translates T-SQL into DuckDB's SQL, and answers the catalog queries that SSMS, VS Code and Power BI ask. It is a single .NET tool, with no database server to install.

DaffyDNA does nothing to a server itself: DaffyDB does the logging in and the reading, and mssql runs your queries. Everything listens only on your own machine.

The warehouse

  • Add sources in a few prompts. A folder of files becomes a database of tables, each file read where it lies, with nothing to import first. Databases mount beside the files, read live.
  • Click a table for its first 100 rows and its columns' types, then Open in SQL Editor to keep going in mssql.
  • Snapshots. Serve a big, slow or production database from a Parquet copy taken now, under the same table names, so your queries don't change.
  • Export to Parquet. One table, or a whole source, to Parquet files to take elsewhere.
  • Open in mssql or Power BI Desktop with the connection filled in.
  • One warehouse per folder, kept in the project, with no passwords in it.

On a SQL Server source, daffydb.run_sqlserver also runs stored procedures and T-SQL within the permissions of the login it was mounted with, so use a read-only login to keep it read-only.

DaffyDB Summary

One page for the whole warehouse: the DaffyDB and DuckDB versions, every mount and schema, where its files and snapshots are, how to connect any client, and the last 50 queries clients sent, each with the DuckDB SQL it became. It also lists every folder's warehouse on this machine, to switch to in a click. Its Power BI section opens Power BI Desktop on the warehouse, with the one step that trips people up called out: choose Database, not Windows, when Power BI asks how to sign in.

Connections: a vault for your logins

SQL Server, PostgreSQL and MySQL logins, and files and folders, saved once and mounted in any folder's warehouse whenever you need them. The SQL Server tab lists your mssql connections too, and shows which are mounted, with Mount, Unmount, Test, New password and Remove beside each. SQL logins and Entra applications (service principals) are supported.

Health

Checks the .NET 10 SDK, DaffyDB, the mssql extension and DaffyDB's certificate, with an Install or Fix button where DaffyDNA can fix it. DaffyDNA installs and updates DaffyDB for you.

Getting started

You need:

  • VS Code 1.105 or later;
  • the .NET 10 SDK. If it's missing, DaffyDNA offers to install it (on Windows, through winget).

Microsoft's SQL Server extension (mssql) is recommended: DaffyDNA uses it to run queries.

  1. Open a folder, trust it, and click the DaffyDNA icon in the activity bar.
  2. In Warehouse, click Start. DaffyDNA offers to install DaffyDB if it isn't installed yet.
  3. Choose a folder of data files, then click a table to see its first 100 rows.

The guide opens by itself the first time DaffyDNA starts, and covers everything else: the ? on each section opens its page.

Your passwords

Passwords you save are kept in VS Code's secret storage, encrypted by the operating system.

  • Where they go: only to DaffyDB, as a parameter of its mount command. Never on a command line.
  • What DaffyDNA never does with them: it never shows or copies a saved password, and never puts one in an mssql or Power BI connection. It never runs anything on a server itself.

The warehouse listens on 127.0.0.1 only, and stops when VS Code closes. In a folder you haven't trusted, DaffyDNA starts nothing.

License

DaffyDNA is MIT-licensed. Copyright (c) 2026 TroBeeOne LLC.

DaffyDB is a separate package that DaffyDNA installs from NuGet. It is MIT-licensed too.

DaffyDNA includes some open-source libraries: markdown-it, and tedious with the Azure libraries it uses. Each keeps its own license, all of them permissive: MIT, ISC, BSD or Apache-2.0. They are listed in THIRD-PARTY-NOTICES.txt.

Protocol and trademarks

DaffyDNA implements nothing of SQL Server. It drives DaffyDB, which implements the Tabular Data Stream protocol from Microsoft's published MS-TDS specification. DaffyDNA is not affiliated with, endorsed by, or sponsored by Microsoft Corporation, and DaffyDB is not affiliated with Babelfish for PostgreSQL. Microsoft, SQL Server, Azure, Entra and Power BI are trademarks of Microsoft Corporation.

Contact

TroBeeOne LLC: trobee.one · LinkedIn

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
  • Your Privacy Choices
  • Consumer Health Privacy
© 2026 Microsoft