skills.rest
richard-gyiko
richard-gyikoCOMMUNITY
Explore all skills by this author →
GitHub

data-wrangler

Execute DuckDB SQL to transform and export data across CSV, Parquet, JSON, Excel, and databases.

3|1|Updated Dec 11, 2025
Search Tags:#analytics#SQL#data transformation#data cleaning#data integration#DuckDB#data export
One-click install
npx skills add https://github.com/richard-gyiko/data-wrangler-plugin --skill data-wrangler
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: data-wrangler
Source: https://github.com/richard-gyiko/data-wrangler-plugin/tree/main/skills/data-wrangler
Command: npx skills add https://github.com/richard-gyiko/data-wrangler-plugin --skill data-wrangler

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires duckdb, polars, pydantic, pyyaml, and includes scripts (resource) components.

What problem does it solve?

This Skill eliminates the manual effort and complexity of data manipulation by enabling powerful SQL-based transformations across diverse data sources. It's ideal for large datasets or complex operations that are beyond simple in-context reasoning, saving you time and reducing errors.

Core Features & Use Cases

  • Multi-Source Data Integration: Seamlessly read and join data from various formats (CSV, Parquet, JSON, Excel) and databases (Postgres, MySQL, SQLite, S3, GCS, Azure, R2).
  • Advanced SQL Transformations: Apply sophisticated DuckDB SQL operations including joins, aggregations, PIVOT/UNPIVOT, sampling, and window functions.
  • Flexible Data Export: Write transformed results to new files (Parquet, CSV, JSON), with options for compression and Hive-style partitioning.
  • Use Case: Imagine you have sales data in a CSV, product information in Parquet, and customer details in a PostgreSQL database. Use this Skill to join these sources, calculate total revenue per customer, and export the result as a partitioned Parquet dataset for your analytics team.

Quick Start

Join orders.parquet with customers.csv and show total orders per customer.

Frequently Asked Questions about data-wrangler

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I join data from multiple formats like CSV, Parquet, and databases in one query?▼

Join multiple data sources by executing DuckDB SQL across CSV, Parquet, JSON, Excel, and databases in a single transformation. The Skill reads all formats, handles multi-source joins, and returns unified results.

Can I transform and export data using SQL without writing custom code?▼

Yes. Execute DuckDB SQL to transform data across diverse sources, then export to CSV, Parquet, or JSON with compression and partitioning options. No custom code needed—configure via YAML.

What data formats does DuckDB SQL support for reading and writing?▼

DuckDB SQL reads and writes CSV, Parquet, JSON, and Excel files, plus connects to Postgres, MySQL, SQLite, S3, GCS, Azure, and R2. Window functions, PIVOT/UNPIVOT, and sampling are supported.

How do I handle sensitive database credentials securely in data transformations?▼

Store credentials in YAML secrets files with environment-variable substitution. The Skill enforces secure credential management and handles Windows shell escaping automatically.

Can I sample large datasets and apply window functions in SQL transformations?▼

Yes. Use DuckDB SQL sampling and window functions within the transformation. Ideal for complex operations on large datasets—reduces manual effort and processing errors.

Related Skills

Explore More Data & Analytics Skills→
cyyehcyyeh
1

data-exploration

Profile datasets to summarize structure, quality, distributions, and anomalies.

Community
Intermediate
starlake-aistarlake-ai
209

parquet2csv

Convert Parquet files into CSV format for batch ETL pipelines.

Official
Intermediate
drewid74drewid74
8

data-engineering

Build idempotent data pipelines converting CSV to Parquet with schema validation.

Community
Advanced
mkurmanmkurman
321

polars

Perform in-memory data manipulation with lazy and eager execution on Apache Arrow.

Community
Advanced