dbt-data-transformation — Claude Code skill: models, tests
A dbt skill for analytics engineering: models, tests, docs, incremental builds, macros and packages. Useful when your warehouse SQL needs structure.
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a claude-codeWhat dbt-data-transformation does
The skill treats dbt as a development framework rather than a pile of scripts. It walks through SELECT-based models, raw sources via {{ source() }}, cross-model references with {{ ref() }}, data quality tests, documentation and exposures. Materializations get their own sections: view, table and incremental with unique_key and on_schema_change.
Project layout is spelled out folder by folder: staging/, intermediate/, marts/, with stg_, int_, fct_ and dim_ prefixes. You get examples of declaring raw tables in sources.yml, attaching unique and not_null tests to columns, and letting dbt resolve the dependency graph. Jinja templating, macros and package management are covered too.
An incremental model breaks if the where clause does not bound the processing window: without a date filter dbt scans the whole table on every run. Check your unique key before the first production run. The text stops inside the incremental section, so orchestration with Airflow, Dagster or dbt Cloud appears only as a bullet in the use-case list, with no detail behind it.
What it helps with
- split SQL transformations into staging, intermediate and marts layers
- declare raw tables and attach column-level tests
- set up an incremental build for a large fact table
- move repeated logic into a Jinja macro
- generate model documentation and a lineage graph
Example prompts for the agent
- Describe the jaffle_shop sources in sources.yml and add unique and not_null tests on id
- Build an incremental fct_orders model with unique_key order_id and an order_date filter
- Lay out models across staging, intermediate and marts with stg_, int_, fct_ and dim_ prefixes
How to install dbt-data-transformation
1Skills CLI (Vercel)
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a claude-code2Manually: put the files in place
~/.claude/skills/dbt-data-transformation/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/dbt-data-transformation/
1Skills CLI (Vercel)
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a codex2Manually: put the files in place
.agents/skills/dbt-data-transformation/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a cursor2Manually: put the files in place
.cursor/skills/dbt-data-transformation/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install manutej/luxor-claude-marketplace dbt-data-transformation2Manually: put the files in place
.github/skills/dbt-data-transformation/SKILL.md1Skills CLI (Vercel)
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a gemini-cli2Platform CLI
gemini skills install https://github.com/manutej/luxor-claude-marketplace --path plugins/luxor-data-engineering/skills/dbt-data-transformationManually: put the files in place
.devin/skills/dbt-data-transformation/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about dbt-data-transformation
How do I install dbt-data-transformation in Claude?
Run npx skills add manutej/luxor-claude-marketplace --skill dbt-data-transformation -a claude-code. The skill lands in ~/.claude/skills/dbt-data-transformation/SKILL.md.
What does dbt-data-transformation do?
It guides dbt development: models, sources, tests, documentation, materializations, macros and packages. Incremental builds and project structure get separate treatment.
How is incremental different from table here?
A table rebuilds fully on every run, while incremental processes only new records after the first build. The example sets unique_key and on_schema_change for incremental.
