DuckDB

When to Use

Activate this skill when the task involves:

• Querying Parquet, CSV, JSON, Iceberg, or Delta files directly with SQL
• Building local analytics pipelines without a database server
• Using DuckDB as an ETL engine for file-to-file transforms (Parquet → Parquet)
• Running DuckDB from Python with DataFrame integration (pandas/Arrow/Polars)
• Reading data from S3, GCS, Azure, or HTTP endpoints
• Using DuckDB as a lightweight alternative to Spark for single-machine workloads
• Replacing pandas for large-file analytical queries

Core Concepts

DuckDB is an embedded analytical database — runs in-process, no server.