# dbt-sugar bootstrap

{% hint style="success" %}
**New in version `0.1.0`**
{% endhint %}

## What is dbt-sugar bootstrap?

The `bootstrap` task is a way to quickly generate model descriptor files (`schema.yml`) for all the models of your dbt project. It can sometimes be daunting if you have not documented many models to get started. **No more excuses with the bootstrap task**. 

Running `bootstrap` is as easy as this:

```
dbt-sugar bootstrap
```

## When to run the bootstrap task?

1. Anytime you want to kick start documentation for any of your models. 
2. Any time you want to make sure that the [audit task](https://bitpicky.gitbook.io/dbt-sugar/commands/dbt-sugar-audit) is 100% accurate. This is because the audit task does not check your database by default so it only knows about coverage for models which have no entry anywhere in `schema.yml` files (see [audit task for more info](https://bitpicky.gitbook.io/dbt-sugar/commands/dbt-sugar-audit)). 

## What does bootstrap really do?

1. First, dbt-sugar will collect all the information about any of the already documented models and keep this safely under its arm! 
2. dbt-sugar will ping your database to get the list of columns of all of your dbt models.
3. If a model's column is already documented, great, if a column is not documented, dbt-sugar will add an entry for it with a place holder that will reads *"No description for this column."* If the column is already documented in another model (see step 1), dbt-sugar will populate it with this description (this is because dbt-sugar is opinionated [see FAQ](https://bitpicky.gitbook.io/dbt-sugar/caveats-and-gotchas/things-to-know-that-we-know-you-should-know#can-dbt-sugar-stop-messing-with-other-model-columns-definitions)).
4. The same happens for models too. If a model is undocumented, dbt-sugar will add an entry for it (with it's columns as described in the steps above).