I built DuckLocal because [your story, one or two sentences, e.g. "I kept opening a notebook or spinning up a database just to answer one question about a CSV"]. DuckDB was already the answer to "how do I query this file fast". What I wanted was an app around it: drop in a file, get a table, write SQL, see the answer. So that's what DuckLocal is. What it does: • Opens CSV, TSV, Parquet, JSON, Excel, whole folders, globs like *.csv, .duckdb files and S3 buckets. Each file becomes a table named after it • Reads files where they are. Nothing is copied or uploaded, and there's no account and no telemetry • SQL editor with autocompletion, formatting and EXPLAIN, plus charts and export to CSV or Parquet • Dashboards as plain-text .dash files, so they diff and review like code, and `ducklocal check` catches a missing column before anyone opens one Built for agents too: The `ducklocal` CLI runs one SQL statement on your local files and prints JSON, with no window and no server. An official agent skill teaches Claude Code and other agents to read the schema before they query. Point your agent at a folder and ask it a question. It's free and open source (Apache-2.0), for Mac (Apple silicon), Windows and Linux. Honest caveats: the Windows build isn't code-signed yet, and Intel Macs need to build from source. I'd love to hear what you'd point it at first, and what's missing for your workflow.

DuckLocal

A DuckDB desktop app for data analysis

A DuckDB desktop app for data analysis