User Guide For UnoPim AI Product Feed

Open AI Product Feed for UnoPim helps businesses create a structured product feed from UnoPim that OpenAI can read.

The module allows you to map attributes, generate feeds in TSV or JSON format, and publish them through a secure token-protected URL.

This guide walks you through every step of configuring feed settings, mapping attributes, generating products feed, and submitting the URL to OpenAI.

• UnoPim 1.0.0 instance with OpenAI Feed package installed.
• For UnoPim installation, make sure your server meets these system requirements.
• An OpenAI / ChatGPT Merchant account for feed submission.

• Generate ChatGPT-compatible product feed from UnoPim in TSV or JSON format.
• Protect feed access using a cryptographically secure token-based URL.
• Map UnoPim product attributes to OpenAI feed fields through a simple admin interface.
• Trigger feed generation manually from the admin panel at any time.
• Generate feed on demand from the command line using the Artisan CLI command.
• Print the authenticated feed URL directly from the CLI using the –show-url flag.
• Monitor the last 5 feed generation attempts with real-time status polling in the dashboard.
• Process large product catalogs in configurable batches to prevent memory errors.
• Set target countries and store country for geographic targeting of AI product listings.

Follow the steps below to install the OpenAIFeed Connector in UnoPim.

Step 1: Extract the Extension

Unzip the extension package and merge the packages folder into the project root directory.

Example structure:

/your-unopim-project
└── packages/

Step 2: Register Service Provider

Open the file: config/app.php

Add the following line under the providers array

Webkul\OpenAIFeed\Providers\OpenAIFeedServiceProvider::class,

Step 3: Register PSR-4 Autoload

Open the project-level composer.json and add the following entry under the psr-4 section:

"Webkul\\OpenAIFeed\\": "packages/Webkul/OpenAIFeed/src"

This ensures Laravel/PHP can autoload the connector classes.

Step 4: Run Setup Commands

Run the following commands from the project root directory.

composer dump-autoload

php artisan migrate

php artisan openai-feed:install

php artisan optimize:clear

Note: For UI theme building, run the npm command inside the package folder: packages/Webkul/OpenAIFeed

npm i && npm run build

You can remove the node_modules after installation.

After the package is installed, the connector appears in the left-hand sidebar of your UnoPim admin panel.

How to Use UnoPim AI Product Feed for openAI search

After the package is installed, the connector appears in the left-hand sidebar of your UnoPim admin panel under the label OpenAI Feed.

Admin Panel → OpenAI Feed

Clicking OpenAI Feed expands two sub-menu items:

• General
• Configuration

Step 1 — Configure General Settings

Go to OpenAI Feed → Configuration. This section controls the core feed behavior. Fill in each field, then click Save Settings in the top-right corner.

Feed Status: Toggle Feed Enabled to activate the feed. When disabled, the feed URL returns no data even if it has been submitted to OpenAI.

Feed Security Token : Click Generate Token to create a cryptographically secure access token. This token is appended to your feed URL as a query parameter.

Format : Select your preferred feed format from the dropdown:

• TSV — Tab Separated Values: Recommended by OpenAI for ChatGPT Search.
• JSON: Available for broader AI platform compatibility.

Channel & Locale : Select the Channel and Locale that match the storefront you want to expose in the feed.

Currency : Enter the ISO 4217 currency code for your pricing (e.g., USD, EUR, GBP). This must match a currency configured on the selected channel.

Store / Brand Information: Add Seller Name and Seller URL.

OpenAI Commerce Flags : The right-side panel contains additional OpenAI-specific settings:

• Eligible for ChatGPT Search: Enable this to allow products to appear in ChatGPT search results.
• Eligible for ChatGPT Checkout: currently our connector does not allow this, hence requires is_eligible_search = false .

Geography & Targeting (Right Panel)

• Target Countries: ISO 3166-1 alpha-2 country codes, comma-separated (e.g., US, GB, CA).
• Store Country: The country where your store is based (e.g., US).

Click Save Settings before moving to the Configuration tab.

Step 2 — Attribute Mapping & Configuration

Go to OpenAI Feed → Configuration. This is where you connect your UnoPim catalog attributes to the fields OpenAI expects in the product feed.

Attribute Mapping

OpenAI requires specific field names in its product feed. Your UnoPim catalog uses your own attribute codes.

The mapping interface bridges the two. For each OpenAI field, select the corresponding UnoPim attribute from the dropdown.

OpenAI Field	Internal Code	Map to UnoPim Attribute	Required
Product Title	title	name / product_name	Required
Description	description	description / short_description	Required
Brand	brand	brand / manufacturer	Required
Price	price	price / retail_price	Required
Image / Gallery Attribute	image	base_image / gallery_images	Required
Sale / Special Price	sale_price	special_price / sale_price	Optional
Weight	weight	weight	Optional
Weight Unit	weight_unit	weight_unit (e.g. kg, lb)	Optional
Color	color	color	Optional

Map at minimum: Product Title, Description, Brand, Price, and Image. Required fields cannot be left empty .

Save Settings after completing all attribute mapping and configuration fields.

Step 3 — Generating Your Product Feed

Once settings and attribute mapping are saved, you are ready to generate the feed. There are three ways to trigger feed generation:

Option A: Manual Generation from Admin Panel

• Go to OpenAI Feed → General in the admin panel.
• Click the Generate Now button.
• The dashboard will show the generation attempt status in real time.
• Once complete, the feed file is ready at your secure token-protected URL.

The dashboard shows the last 5 generation attempts with timestamps and status. Use this to confirm success or identify failures.

Option C: CLI Generation

Generate the feed on demand from the command line:

php artisan openai-feed:generate

To generate the feed and print the authenticated public URL in the terminal:

php artisan openai-feed:generate –show-url

Copy the output URL — this is the URL you submit to OpenAI.

Step 4 — Retrieving Your Feed URL

After the feed is generated, you can retrieve the secure feed URL in two ways:

From the Admin Panel

• Go to OpenAI Feed → General.
• After a successful generation, a Download Feed button and a Copy URL button appear in the dashboard area.
• Click Copy URL to copy the token-authenticated feed URL to your clipboard.

From the CLI

Run the following command from your server:

php artisan openai-feed:generate –show-url

The full URL including the security token is printed to the terminal.

The feed URL format is: https://yourdomain.com/openai-feed?token=YOUR_SECURE_TOKEN — the token parameter authenticates the request to OpenAI.

Step 5 — Submitting the Feed URL to OpenAI

With your feed URL ready, submit it to OpenAI to make your products visible to OpenAI Search.

• Log in to your OpenAI Merchant account.
• Navigate to the Product Feed or Shopping settings section.
• Paste your token-authenticated feed URL into the feed submission field.
• Select TSV as the feed format (or JSON if that is what you generated).
• Save and submit the feed for OpenAI review.

OpenAI will crawl and index your feed. Once approved, your products will begin appearing in ChatGPT search results for relevant queries.

Step 6 — Monitoring Feed Status

The admin dashboard logs every feed generation attempt. Navigate to OpenAI Feed → General to view the status panel.

The dashboard shows the last 5 attempts. Each row displays the generation method (manual, cron, CLI), timestamp, and result status.

Step 7 – Preview Feed

From the Admin Panel

• Go to OpenAI Feed → General.
• Click on Preview Feed .

This was all about UnoPim AI Product Feed Connector. I hope you have got the idea of our module. Thanks for taking the time to read this blog.

And also, please don’t forget to share your feedback and suggestions in the comment box given below.

Moreover, If you have any queries or suggestions, then feel free to add a ticket at our HelpDesk system or reach out to us at support@webkul.com.