> For the complete documentation index, see [llms.txt](https://usermanual.ergonode.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://usermanual.ergonode.com/products/catalog/ai-actions/generate-descriptions.md).

# Generate descriptions

{% hint style="info" %}
You can choose only one attribute and one language at a time.
{% endhint %}

{% hint style="info" %}
Read [this article](https://usermanual.ergonode.com/tutorials-and-video-clips/how-to-write-effective-prompts) to learn how to write effective prompts.
{% endhint %}

{% hint style="info" %}
You can use [variables](https://usermanual.ergonode.com/product-design/attributes/attribute-variables) in the prompt window. To do so, type #, and a list of available attributes will be displayed.
{% endhint %}

{% hint style="warning" %}
Keep in mind that AI-generated content may contain errors. Review the generated values carefully.
{% endhint %}

{% hint style="info" %}
Supercharge product descriptions and SEO using AI prompts - free guide available in [Polish](https://www.ergonode.com/pl/zasoby/e-booki/jak-pisac-prompty-pim) and [English](https://www.ergonode.com/resources/e-books/how-to-write-prompts-for-ai-pim).
{% endhint %}

To open this action, click on **AI Actions** in the bar above the grid and choose **Generate descriptions**.

First, specify the textarea attribute and the language in which the content will be generated. A list of applicable attributes to choose from is available in the **Attribute** dropdown. A list of applicable languages to choose from is available in the **Language** dropdown.

Next, type your prompt to generate the descriptions. List the specific attributes you want to include in the description. Write a clear and concise prompt that includes the attributes and additional instructions.

To add a variable, type # and pick the attribute from the list: the variable is then highlighted in the prompt. An attribute code you type in full without picking it is sent as plain text, so the AI does not get the product's value.

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2FqpQjS8D91UZzuCwkD8Nf%2Fimage.png?alt=media&#x26;token=7cdc289a-c20f-4cb0-8499-59ef0434eb1a" alt="" width="563"><figcaption></figcaption></figure>

Before you run the prompt, choose the **AI Model**. Each model in the list shows a multiplier next to its name (x1, x2, x20): it is the multiplier of the AI credits the model uses. For example, a x2 model uses twice as many credits as a x1 model. You can also use the **Require values for attribute variables** toggle: when it is on, products with no value in an attribute used as a variable are not processed and are reported as errors. When it is off, these products are processed too.

Click the **Try out** button to generate value for the first product only, to test the prompt, before applying it to all products. The result shows on the right: the sample product, your prompt with the variables replaced by its values, and the generated text. Nothing is saved yet.

{% hint style="warning" %}
Even test runs of the prompt will count toward the monthly AI usage limit for your instance. The **Try out** and **Generate** buttons show the estimated number of AI credits they use.
{% endhint %}

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2F9BVQvZwJVGfDtFvzWqHl%2Fimage.png?alt=media&#x26;token=93b6115e-0774-44d8-ad7c-034dc006a77d" alt="" width="563"><figcaption></figcaption></figure>

Click the **Generate** button to generate value for all selected products at once. A confirmation window will show up. Click **Start batch action** to start or **Cancel** to go back to the form.

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2FFc6TIr2J3FqWw1UfA1C2%2Fimage.png?alt=media&#x26;token=9ab99941-73b1-4a64-8d3c-53dbac4a1b4f" alt="" width="563"><figcaption></figcaption></figure>

The progress shows in the same window. You can close it: the action continues, and you can follow it in the [Action center](https://usermanual.ergonode.com/dashboard/dashboard/user-menu-and-action-center/action-center). You can stop the action from the ***three dots*** menu with **Stop action**.

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2FpQISPx4KkHBfbuGtcCt6%2Fimage.png?alt=media&#x26;token=86e7d1d6-7b4e-426b-87d6-c8502e738d69" alt="" width="563"><figcaption></figcaption></figure>

If an error occurs, you can choose **See details** from the ***three dots*** menu or click the **See details** button.

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2FpxTiFeW9YnWOMiUfzahy%2Fimage.png?alt=media&#x26;token=f6136d91-5d58-4d20-9264-0dfec6ae01ad" alt="" width="563"><figcaption></figcaption></figure>

The action details list every product with its status and error message, for example, a product with no value in an attribute used as a variable.

<figure><img src="https://171427267-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FM9BbYBbDPiJapb75xJ2k%2Fuploads%2FqUSUor07yROMNE7EwTLW%2Fimage.png?alt=media&#x26;token=ac592ce7-d3e0-4520-a262-bafd3e87e105" alt="" width="563"><figcaption></figcaption></figure>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://usermanual.ergonode.com/products/catalog/ai-actions/generate-descriptions.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.