dlt (Data Load Tool)

Build Python ELT pipelines with dlt. Loads data from any API, file, or database into warehouses and lakes with automatic schema inference, incremental loading, and built-in data contracts — the requests library for data pipelines.

This skill makes Claude a dlt expert. It scaffolds pipelines that extract from REST APIs, files, and databases, then load into DuckDB, BigQuery, Snowflake, Postgres, and more. Covers automatic schema inference and evolution, incremental and merge write dispositions, resource and source decorators, state management, and deployment to Airflow, GitHub Actions, or dlt+ — with data-contract enforcement at the boundary.

When to use

Use when building a Python ELT pipeline, ingesting from a REST API into a warehouse, setting up incremental loading, or handling schema evolution with dlt.

Examples

Load a REST API

Ingest into a warehouse

Build a dlt pipeline that pulls the GitHub issues API into DuckDB with incremental loading on updated_at

Schema evolution

Handle changing sources

Set up a dlt source with merge write disposition and let it auto-evolve the schema when new columns appear

Deploy to Airflow

Productionize the pipeline

Wrap my dlt pipeline as an Airflow DAG that runs hourly and loads into BigQuery