Skip to content

Install

macOS and Linux (x86_64 and ARM):

Terminal window
curl -fsSL https://getdre.com/install.sh | sh

This puts dre in ~/.local/bin, after checking the download against the release’s SHA256SUMS. DRE_INSTALL_DIR picks another folder and DRE_VERSION a release (default: the newest, pre-releases included):

Terminal window
curl -fsSL https://getdre.com/install.sh | DRE_VERSION=v0.1.2 DRE_INSTALL_DIR=/usr/local/bin sh

The same line works in a Databricks job (a cluster init script or a %sh cell), a CI runner or a container build. On Windows, download dre-<version>-windows-x86_64.zip (or -aarch64.zip) from Releases and put dre.exe on your PATH.

From 0.1.0 on:

Terminal window
brew install get-dre/tap/dre # macOS and Linux
Terminal window
scoop bucket add get-dre https://github.com/get-dre/scoop-bucket # Windows
scoop install get-dre/dre

The dre-cli package holds dre (Linux x86_64 and aarch64, macOS, Windows; Python 3.8+):

Terminal window
pip install dre-cli # or: uv tool install dre-cli
dre --version

Each release attaches the wheels (dre_cli-*.whl), which install the same way with pip install <wheel URL> before the version reaches PyPI. From Python, dre_cli.run(["run", "-s", "daily"]) runs dre and returns the finished process. In a Databricks job (serverless included), add dre-cli to the job’s environment dependencies and run dre from a script or notebook.

Terminal window
cargo install dre-cli --locked # builds `dre` from crates.io

A pre-release installs only when named: cargo install dre-cli --locked --version 0.2.0-rc.1.