vs-quack
vs-quack is a lightweight, fast, and convenient DuckDB-powered data exploration and processing extension for Visual Studio Code.
Run SQL queries, preview data files, inspect variables during debugging, and view query history — all backed by DuckDB.
Features
If you already use DuckDB with a .duckdbrc or other config file, load your settings via:
vs-quack: Set DuckDb Settings (Command Palette)
Why this matters:
This loads your DuckDB configuration (e.g., httpfs, AWS credentials, S3 endpoints, access keys).
✔ You must do this if you want to query files stored in S3.
Once configured, vs-quack can query S3 paths directly, such as:
SELECT * FROM 's3://my-bucket/data.parquet';
You can also perform cross-file and cross-format operations, like joining a CSV with a Parquet file:
SELECT * FROM 'local_file.csv' AS a
JOIN 's3://bucket/data.parquet' AS b
ON a.id = b.id;
For more information about duckdb's cloud integrations visit:
⚡ Run DuckDB Queries
You can execute queries in multiple ways — pick whichever suits your workflow:
1️⃣ Run a query from a prompt
Open the Command Palette →
vs-quack: Run DuckDb Query
Type your SQL query and select between execute, or export in different formats.
💡 Perfect for ad-hoc queries or testing small snippets, as well as data importing.
2️⃣ Run selected text as a query in the editor
Highlight the SQL text in your editor → Right-click →
vs-quack: Run Selected Text as Query
🔹 Great for working within scripts or notebooks.
3️⃣ Run a query on a file via Explorer
Right-click a file in the Explorer →
vs-quack: Run Query on Explorer
🗄️ Instantly query CSV, Parquet, JSON, or any DuckDB-supported file.
4️⃣ Quick preview with “Choose File”
Use vs-quack: Choose File → select a file → opens a blank query ready to run.
👀 Ideal for inspecting data files without writing queries first.
📝 Interactive Notebook Support
Create a .duckdbnb file and run queries in a VS Code notebook powered by DuckDB:
- Each cell can contain SQL queries (default cell type: sql)
- Execute individual cells or the entire notebook
- View query results inline, with full support for large datasets
- Perfect for exploratory data analysis across multiple files
💡 Notebooks make it easy to mix queries, visualizations, and notes in one place.
🐛 Evaluate SQL from Debug Variables
Evaluate SQL stored in variables during debugging:
Right-click → vs-quack: Evaluate Variable
📁 Preview CSV / Parquet / JSON Files
Use vs-quack: Choose File to select a data file and open it with an empty query.
Useful for quickly previewing CSV, Parquet, JSON, and other structured formats.
📜 Query History
View your last 50 executed queries, filter them, and reopen them in the editor with:
vs-quack: Show Query History
Requirements
- Visual Studio Code v1.97.0 or higher
Known Issues
- Currently, only macOS (darwin-arm64) and Windows (win32-x64) are supported; Linux is not yet supported.
- No internal DuckDB limit on database size (when creating tables)
- No memory or CPU limits enforced by the extension
→ You can control these through your DuckDB settings file
Release Notes
2.0.141
- Updated to DuckDB 1.4.1
- Results view styling updated to match VS Code UI
- Only one results panel remains open at a time
- Limited results to top 200 + bottom 200 rows
- Disabled auto-open when running queries from editor selection
2.1.141
- Added evaluation of query variables during debugging
- Added “Choose File” feature
- Moved row limitation into the reader (top 500 rows)
3.0.141 - 3.1.141
- Visual improvements (warnings, executed query, progress indicator)
- Added query execution time to results window
- Increased result preview limit to 999 rows
- Introduced notebook support — interactive SQL notebooks with inline results
4.0.141
Added Windows support through native node modules.
