Lightweight, Power BI–inspired data visualizer powered by DuckDB-WASM and Apache ECharts.
PowerDuck is a local-first data visualizer for VS Code. Open data files from Explorer, build visuals with drag-and-drop fields, and inspect or edit the SQL behind each chart.
Features
- Open local
.parquet,.csv,.tsv,.json,.jsonl, and.ndjsonfiles directly from VS Code Explorer. - Build visuals with drag-and-drop field wells (Axis, Legend, Values, Columns, Filters).
- Use multiple visual types: bar/column variants, line, area, scatter, pie, doughnut, table, KPI card.
- Inspect and edit generated SQL, then re-run with
Ctrl+Enter. - Keep data local: all analysis runs in DuckDB-WASM inside the webview.
Getting Started
- Install the extension.
- In VS Code Explorer, right-click a supported data file.
- Choose Open in Data Visualizer (PowerDuck).
- After initialization, create visuals and optionally edit SQL.
Supported Formats
| Extension | DuckDB reader |
|---|---|
.parquet |
read_parquet |
.csv, .tsv |
read_csv_auto |
.json, .jsonl, .ndjson |
read_json_auto |
Privacy & Security
- Query execution is client-side in DuckDB-WASM.
- Local files are streamed from extension host to webview as
Uint8Arraychunks and are not uploaded by the extension. - The webview uses a strict Content Security Policy.
- On first non-CSV use, DuckDB-WASM may fetch a signed format extension
module from
https://extensions.duckdb.org(no user data upload).
Tech Stack
Support
- Issues: GitHub Issues
- Repository: atliuhui/powerduck
License
MIT. See LICENSE.
PowerDuck is an independent open-source project. Power BI is a trademark of Microsoft Corporation; this project is not affiliated with or endorsed by Microsoft.
Contributing? See DEVELOPMENT.md.