> ## Documentation Index
> Fetch the complete documentation index at: https://fromhazel.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# How to Run Recipes in Hazel

> Run a recipe with a simple slash command or let Hazel automatically invoke the right analytics workflow based on your natural language question.

There are two ways to put a recipe to work: call it directly with a slash command, or let Hazel decide when to use it.

## Slash commands

Every installed recipe has a slash command. To run it, type `/` in the chat input, pick the recipe from the palette, and send the message.

<Steps>
  <Step title="Type /">
    A palette appears above the input showing every installed, enabled recipe you have access to.
  </Step>

  <Step title="Filter by name or command">
    Keep typing to narrow the list — `/ltv`, `/cohort`, `/brand`. Arrow keys navigate, Enter selects.
  </Step>

  <Step title="Add context and send">
    The slash command tells Hazel which recipe to run. Anything else in the message gives it the specifics — date range, segment, product line, whatever's relevant.
  </Step>
</Steps>

For example:

```
/ltv for subscribers acquired in Q1, 12-month horizon
```

Hazel loads the full recipe instructions before it starts thinking, so the analysis follows the recipe's rules from the first step.

## Automatic invocation

You don't always need to know a recipe exists. Hazel sees the name and description of every enabled recipe and can pull in the right one on its own when your question matches.

Ask something like *"What's our best customer cohort?"* and Hazel may decide to run your LTV recipe without being told. When it does, you'll see the recipe referenced in its response so you know which playbook it used.

<Tip>
  This is why **descriptions matter**. A clear, specific description helps Hazel match the recipe to the right questions. Vague descriptions mean the recipe gets picked rarely or at the wrong moments.
</Tip>

## Enabled vs. disabled

Every recipe has an on/off switch on its detail page.

* **Enabled** — visible in the slash palette and considered for automatic invocation
* **Disabled** — stays installed but is hidden from Hazel entirely

Disable recipes you've outgrown, recipes that are still being tuned, or recipes you only want used when explicitly invoked. Disabling is reversible; deleting is not.

## Who sees what

Which recipes show up in your slash palette depends on scope:

* Personal recipes — only yours appear for you
* Organization recipes — appear for every teammate in the workspace

If you don't see a teammate's recipe, it's either Personal-scoped to them or disabled.

***

**Next up:** [Create your own recipe](/docs/recipes/creating-recipes) from the cookbook or from scratch.