Duckdb

This gives Claude the ability to run SQL queries directly on your data files without spinning up a database server. Point it at CSVs, Parquet files, or JSON and write analytical queries with joins, window functions, and aggregations. It can also query pandas and Polars DataFrames in place, which is genuinely useful when you're neck-deep in a Jupyter notebook and don't want to wrestle with method chaining. The performance is solid for medium to large datasets since it's built for OLAP workloads. Think of it as SQLite's analytical cousin. Best when you need complex SQL on local files or want to join data across different formats without loading everything into memory first.