Getting Started

Prerequisites

Lakecode requires Node.js 18 or later. Check your version with:

node --version

If you don't have Node.js, install it from nodejs.org or with your package manager of choice.

Install

npm install -g lakecode

Verify it works:

lakecode --version

Login

Authenticate with your Google account. This grants Lakecode access to manage your credits and proxy AI requests on your behalf.

lakecode login

This opens your default browser for Google OAuth sign-in. Once authorized, a license key is saved locally and used for all subsequent requests.

If you are on a headless server or SSH session where a browser is not available, use the device flow instead:

lakecode login --device

This prints a URL and a code. Open the URL on any device, enter the code, and sign in with Google. The CLI will detect the authorization and save your key automatically.

First command

Run your first evidence pack against any table in your Databricks or Snowflake workspace:

lakecode /prove my_catalog.my_schema.my_table

This generates a full health assessment: schema snapshot, row counts, null analysis, freshness, duplicates, numeric stats, and sample rows.

Workspace profiling

On first run, Lakecode automatically profiles your Databricks or Snowflake workspace. It discovers catalogs, schemas, tables, and warehouses available to your account. This profile is cached locally so future commands have instant context about your environment. You do not need to configure anything — it happens automatically.