# Prompts

## Overview

The prompts are a necessary component of the AI Discovery Experience setup. On their own they do not serve any purpose, but once paired with a user query, they help provide better context to the AI agent to process that user query. 

Please visit the [AI Discovery page](/platform-documentation/catalog-2.0-docs/project-data/experiences/ai-discovery.md) for more detailed information on how to use the Prompts and what kind of instructions you would want to write in them.

## Creating Prompts

While on the Prompts page, please click on the Create Prompt button in the top right corner of the page to start creating a new Prompt entry.

<figure><img src="/files/jWRMUlamzjJkZ2np4lyn" alt=""><figcaption><p>Creating a Prompt</p></figcaption></figure>

### Title

The title of the prompt as it will appear in the platform UI when selecting it on the Prompt page or in the Rephrase and Confidence Prompt dropdowns on the AI Discovery page.

### Custom ID

Assign your own custom ID to the prompt, which can be used instead of the standard ID assigned by the platform to reference the prompt with the API.

### Description

You can provide a description for your prompt that is only intended to be used for your own documentation purposes. This could help to give yourself or your team more context for why you created the prompt the way you did, and how you plan to use it.

### Prompt Text

This text needs to be filled out in order to truly make the Prompt functional. This is the area where you can type instructions to the AI agent on how to process certain types of input. There are currently only two uses for Prompts, as described on the AI Discovery page.

* Rephrase Prompts - a set of instructions to the AI agent on how to process a user search query for a product.
* Confidence Prompts - a set of instructions to the AI agent on how to "verbally" present the query results to the end user. Instruct it on how to write compelling reasons why the presented results would make great purchase choices that fit the user query.

{% hint style="success" %}
**These instructions for the AI agent are key to providing the end users with great search query results.**&#x20;

[Visit these instructions from OpenAI for more information on great prompt engineering](https://help.openai.com/en/articles/10032626-prompt-engineering-best-practices-for-chatgpt)
{% endhint %}


---

# Agent Instructions: Querying This Documentation

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

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://community.threekit.com/platform-documentation/catalog-2.0-docs/project-data/experiences/prompts.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.