databricks/cli

What it is

Databricks CLI is the official command-line client for working with Databricks workspaces. It ships as a dependency-free binary that can be placed in any directory, so it fits scripts, CI jobs, and local terminals that need to call the Databricks platform without a separate runtime. The command surface includes workspace operations and bundles; the project documents commands and flags in the linked Databricks documentation documentation.

Getting started

1. Install the CLI. Downloads are published on the releases page, and installation choices are documented in the Databricks CLI install guide. Homebrew users can follow the Homebrew tap README; container users can pull ghcr.io/databricks/cli:latest from the published container image.
2. Authenticate using Databricks Unified Authentication; the supported methods are described in the authentication guide.
3. Run a command to confirm the client works. With the container image, the README gives this example:

docker run -e DATABRICKS_HOST=$YOUR_HOST_URL -e DATABRICKS_TOKEN=$YOUR_TOKEN ghcr.io/databricks/cli:latest current-user me

Replace $YOUR_HOST_URL and $YOUR_TOKEN with your workspace URL and token before running it air-gapped environment documentation.

Stability and versioning

Stable commands and flags are guaranteed not to break within a major version, while commands marked Beta or Private Preview in --help, and commands under databricks experimental, may change in any minor release. Releases follow Semantic Versioning SemVer, and security fixes are described in SECURITY.md. The project states that it does not currently offer a broader long-term support commitment.

Usage constraints

The CLI is a command-line tool, not a hosted service, and it requires Databricks authentication credentials for the workspace it targets. Use of the CLI is governed by the Databricks License and the Databricks Privacy Notice; the repository does not use a standard OSI license label. There is no separate long-term support path beyond the security policy.

HotRepo's take

On 2026-11-12 the project recorded 429 stars and 269 forks, with a daily gain of 2 stars and a best position of 5 on the Go daily list. That snapshot is the only record available here, and the rank is specific to that language list and date rather than an all-language daily rank. The notable signal is the fork-to-star ratio: this is a client many people pull into their own automation rather than a fast-growing end-user tool, which fits its role as official Databricks tooling.