data-skills

A collection of Claude Code skills for data engineering tasks.

Overview

This repository contains expert assistant skills designed to help with data engineering workflows. Each skill provides specialized knowledge, templates, and best practices for specific data tools and patterns.

Available Skills

Skill	Description	Docs
dlt-skill	Data pipelines with dlt (data load tool)	Guide
dlt-dagster	Run dlt pipelines in Dagster (Component or Pythonic @dlt_assets)	—
uv	Python dependencies, venvs, and scripts with uv	Guide

Installation

Install skills using the Claude Code CLI:

npx skills add untitled-data-company/data-skills --skill <skill-name>

Replace <skill-name> with the skill directory name (e.g. dlt-skill, dlt-dagster, uv). See the Available Skills table above.

Verify Installation

After installation, the skill should appear when you run /skills in Claude Code.

Repository Structure

data-skills/
├── skills/
│   ├── dlt-skill/       # Data pipelines (dlt)
│   │   ├── SKILL.md
│   │   ├── assets/
│   │   ├── references/
│   │   └── scripts/
│   ├── dlt-dagster/     # dlt pipelines in Dagster
│   │   ├── SKILL.md
│   │   ├── assets/templates/
│   │   └── references/
│   └── uv/              # Python venvs & deps (uv)
│       ├── SKILL.md
│       └── references/
├── docs/                # User documentation
│   ├── dlt-skill.md
│   └── uv.md
├── README.md
└── CHANGELOG.md

Contributing

To add a new skill:

1. Create a new directory under skills/
2. Add a SKILL.md with the skill instructions (frontmatter: name, description)
3. Include any templates, references, or scripts needed
4. Add user documentation in docs/<skill-name>.md
5. Add an entry to the "Available Skills" table above

License

Skills in this repository are provided as-is for use with Claude and other skill-enabled LLMs. Maintained by Untitled Data Company.