# Manage data structures via Snowplow CLI
> Use the Snowplow CLI data-structures command to generate, download, validate, and publish data structures with git-ops workflows and JSON Schema validation.
> Source: https://docs.snowplow.io/docs/api-reference/snowplow-cli/data-structures/
> Documentation index: https://docs.snowplow.io/llms.txt

- Cloud: available
- Private Managed Cloud: available
- Self-Hosted: not available
- Add-on: not available

Managing data structures and tracking plans requires a Snowplow CDI (Console) account.

Find out more about [Snowplow CDI](/docs/cdi/) and [product availability](/docs/get-started/feature-comparison/), or contact [Support](https://support.snowplow.io) to upgrade.

The `data-structures` subcommand of [Snowplow CLI](/docs/api-reference/snowplow-cli/) provides a collection of functionality to ease the integration of custom development and publishing workflows.

## Snowplow CLI Prerequisites

Installed and configured [Snowplow CLI](/docs/api-reference/snowplow-cli/)

## Available commands

### Creating data structures

```bash
snowplow-cli ds generate login_click ./folder-name
```

Will create a minimal data structure template in a new file `./folder-name/login_click.yaml`. Note that you will need to add a vendor name to the template before it will pass validation. Alternatively supply a vendor at creation time with the `--vendor com.acme` flag.

### Downloading data structures

```bash
snowplow-cli ds download
```

This command will retrieve all organization data structures. By default it will create a folder named `data-structures` in the current working directory to put them in. It uses a combination of vendor and name to further break things down.

Given a data structure with `vendor: com.acme` and `name: link_click` and assuming the default format of yaml the resulting folder structure will be `./data-structures/com.acme/link_click.yaml`.

> **Note: Drafts not included**
>
> The CLI download command only retrieves data structures that have been deployed to at least the development environment. **Draft data structures** that haven't been deployed yet will not be included in the download.

### Validating data structures

```bash
snowplow-cli ds validate ./folder-name
```

This command will find all files under `./folder-name` (if omitted then `./data-structures`) and attempt to validate them using Snowplow Console. It will assert the following

1. Is each file a valid format (yaml/json) with expected fields
2. Does the schema in the file conform to [snowplow expectations](/docs/fundamentals/schemas/#self-describing-json-schema-anatomy)
3. Given the organization's [loading configuration](/docs/destinations/warehouses-lakes/) will any schema version number choices have a potentially negative effect on data loading

If any validations fail the command will report the problems to stdout and exit with status code 1.

### Publishing data structures

```bash
snowplow-cli ds publish dev ./folder-name
```

This command will find all files under `./folder-name` (if omitted then `./data-structures`) and attempt to publish them to Snowplow Console in the environment provided (`dev` or `prod`).

Publishing to `dev` will also cause data structures to be validated with the `validate` command before upload. Publishing to `prod` will not validate but requires all data structures referenced to be present on `dev`.