duckdb

CLI

In-process analytical SQL database engine.

Run SQL analytics on CSV, Parquet, and JSON files without a server

2 curated recipes · not yet executed by the harness#sql #analytics #olap #parquet #csv

Recipes

Curated from documentation and usage. Percentages are estimates, not measurements.

Query CSV with SQL

~98% est.

duckdb -c "SELECT count(*), avg(price) FROM 'data.csv'"

Convert Parquet to CSV

~97% est.

duckdb -c "COPY (SELECT * FROM 'data.parquet') TO 'output.csv' (HEADER, DELIMITER ',')"

Homebrewbrew install duckdb

When to use: Analysing large datasets with SQL directly from files
Quick examples: duckdb -c "SELECT * FROM 'data.csv' LIMIT 10"
duckdb -c "COPY (SELECT * FROM 'data.parquet') TO 'out.csv' (HEADER)"

curl -s https://clihub.com/api/tools/duckdb | jq '.agentHints.provenRecipes[] | {task, command, verified}'

Or call get_tool with slug duckdb on the MCP server at https://clihub.com/mcp.