> ## Documentation Index
> Fetch the complete documentation index at: https://docs.t3planet.de/en/latest/llms.txt
> Use this file to discover all available pages before exploring further.
# Feature Guide
> Feature guide for T3AC (EXT:ns_t3ac) chatbot, training, and analytics.
## Dashboard
* View, activate, deactivate, and manage all created chatbots.
* Create new chatbot instances.
## Configuration
While creating or editing a chatbot, you can configure:
* Title (e.g., AI Chatbot)
* Short Description
* Bubble Message (example: “Hey there, How can I help you?”)
* Welcome Message (example: “Hi, how can I help you?”)
* Chatbot Language
* Chatbot Instructions
## Customization
Customize the appearance and behavior of the chatbot:
* Upload Logo
* Upload Avatar
* Avatar Size
* Chatbot Color
* Show/Hide Logo
* Show/Hide Date & Time
* Transparent Trigger Option
* Chatbot Position (e.g., bottom-right)
* Widget Bottom Space (In Pixel Min: 15px, Max: 200px)
## Training
T3AC offers multiple methods to train your chatbot for better and specific responses:
**Meaning of Status Labels:**
* **Untrained** means the Chatbot has not been trained on the data yet.
* **Trained** means the Chatbot has already been trained and has learned from the data.
**Typo3 Sitemap:**
This feature allows the chatbot to automatically fetch URLs from your TYPO3 website’s sitemap, scan the pages, and extract content for AI training — helping you keep the chatbot’s knowledge base up to date and well-trained.
**Website Training:**
* Website Mode: Enter a website URL to fetch and train from all available pages.
* Single URL Mode: Enter a single webpage URL to train specifically.
**PDF Training:**
* Upload a PDF file and allow the chatbot to learn from its content.
**Text Training:**
* Manually add a custom Title and corresponding Text for specific training.
**Q\&A Training:**
* Add Question & Answer pairs to create a focused FAQ-based training set.
After feeding the data, simply select and click *Train GPT*.
## Embed
If you’d like to use this chatbot on another domain, follow these simple steps:
* Go to tab Embed
* Click the Copy button to copy the embed code.
* Paste the code inside the **``** tag of your website.
* Allow the specific domain in the .htaccess file of the website where the chatbot is created to enable proper access.
## Chat Logs
* Access all conversations recorded by the chatbot.
* Monitor interactions, review conversations, and improve chatbot training based on real user inputs.
## Multilanguage
Configure your chatbot to support multiple languages directly from the **Configuration** tab by selecting the default language from the dropdown menu. Only the languages that have already been added to your domain will be available for selection.
## Scheduler
Use the **Scheduler** under the **Website** section of the **Training** tab to automate training tasks on large website datasets—ideal for frequent content updates or large-scale web crawling.
## Sitemap Crwaling
* To run the scheduler via CLI, use the command (1 is the ID of the scheduler):
```bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
vendor/bin/typo3 nst3ac:crawl:sitemap "https://v13-composer.ddev.site/sitemap.xml" 1
```
* This scheduler allows you to crawl the content of all URLs defined in the sitemap.
* When executed via the CLI, it displays the crawl progress.
* After execution, a summary of successfully processed and failed URLs is shown.
## Chatbot visibility
**Show Chatbot on Specific Pages**
* Enter the Page IDs where the chatbot should be visible (e.g., 10, 22, 35).
**Hide Chatbot on Specific Pages**
* Enter the Page IDs where the chatbot should be hidden, even if it is enabled globally (e.g., 45, 60, 72).
## Basic Authentication Support
If you want to crawl URLs protected by htaccess, simply enable the Basic Authentication checkbox and enter the htaccess username and password.
## Configurable Database Chunking for various types of Dataset Processing
**Overview**
* This feature provides configurable database chunking to efficiently process large datasets while maintaining optimal performance and stability. Instead of loading all records in a single query, data is fetched and processed in smaller chunks, reducing memory usage and preventing execution timeouts.
* The chunk size is configurable via the extension settings, allowing administrators to adjust it according to server capacity and dataset size.
**Default Configuration**
* By default, database chunking is enabled with a chunk size of 1000 records.
* If required, this value can be modified in the extension settings to better suit the execution environment.
