DuckDB SQL

Generates and refines DuckDB SQL queries for data analysis using local database, CSV, and Parquet files.

SkillDatabase ManagementSqlAnalytics Warehouse

The source repository doesn't declare a license. Check its terms before reusing the code.

Key features

  • Operates in a 'display-only' mode, showing queries for review before execution
  • Follows best practices for DuckDB, including ATTACH statements and relative pathing
  • Analyzes data to automatically detect and document potential enum columns
  • Creates a local asset cache with schemas and a data dictionary for fast query generation
  • Generates DuckDB-compliant SQL for .ddb, .csv, and .parquet files

Use cases

  • Quickly generating SQL to explore and analyze local data files without remembering exact schemas.
  • Creating a persistent, documented data dictionary for a project's local data sources.
  • Refactoring, improving, or debugging existing DuckDB SQL queries for better performance.

FAQ

How does this skill improve my workflow?

It accelerates your data analysis by creating and maintaining a data dictionary and schema files. This 'asset cache' provides Claude with deep context, leading to faster, more accurate queries. Its 'display-only' default also ensures you can review all SQL before execution for safety.

What does the DuckDB SQL skill do?

This skill acts as a DuckDB expert, generating and refining SQL queries for data analysis on your local database (.ddb), CSV, and Parquet files. It automatically creates a local asset cache, including schemas and a data dictionary, to make query generation fast and accurate.

What are its main capabilities?

The skill generates DuckDB-compliant SQL, creates a reusable asset cache with schemas and a data dictionary, operates in a 'display-first' mode for query review, automatically documents potential enum columns, and follows DuckDB best practices like using ATTACH statements.

When should I use this skill?

Use this skill when you need to explore local datasets with DuckDB. It's perfect for quickly generating queries from plain English, understanding the schema of new files, or getting expert feedback to improve your existing SQL code.