databricks-workspace-explorer

A VS Code extension that mounts your Databricks Workspace as dbws:/ so you can browse and edit workspace files from the Explorer.

What it does

Registers a filesystem provider at:

• dbws:/

So you can open Databricks workspace paths directly in VS Code Explorer.

Authentication

This extension uses the Databricks JS SDK for authentication, which supports multiple methods:

1. Databricks CLI profile (recommended) — set up once with databricks configure, then set the VS Code setting databricksWorkspace.profile (leave it empty to use the Databricks CLI default profile).
2. Environment variables — DATABRICKS_HOST + DATABRICKS_TOKEN.
3. OAuth (databricks-cli) — run databricks auth login first.
4. Azure CLI — for Azure Databricks workspaces.

No token is ever stored in VS Code settings. Credentials are managed by the standard ~/.databrickscfg file or the Databricks CLI's OAuth token cache.

Quick start

# Install Databricks CLI and authenticate
databricks auth login --host https://<your-workspace>.cloud.databricks.com

Then set the workspace host in VS Code settings (optional — the SDK reads it from your .databrickscfg):

{
  "databricksWorkspace.host": "https://<your-workspace>.cloud.databricks.com"
}

Run the extension locally

npm install
npm run compile
npm run test:all

Then press F5 in VS Code (Extension Development Host).

Open Databricks in Explorer

In the Extension Development Host:

1. Command Palette → Databricks: Open Workspace Explorer
2. VS Code opens dbws:/ as a folder
3. Browse files/folders in the Explorer panel

Commands

Available from the Command Palette:

• Databricks: Open Workspace Explorer — opens the dbws:/ workspace root in Explorer.
• Databricks: Open Workspace Path — opens a specific Databricks workspace file or folder by path.
• Databricks: Refresh Workspace — refreshes the Databricks workspace view.
• Databricks: Show Active Auth/Profile — shows the active host, profile, and optional Databricks CLI path.

Notes

• Supports create/write/delete for workspace files and notebooks.
• Notebook files are shown as .ipynb in Explorer and use JUPYTER import/export.
• Script/text files use SOURCE import and AUTO export.
• Directory rename is not supported yet.