This commit is contained in:
Rodrigo Rodriguez 2024-09-04 16:58:16 -03:00
parent d8fa41515f
commit 15cff0379e
2 changed files with 100 additions and 1 deletions

1
.gitignore vendored
View file

@ -3,3 +3,4 @@ deploy.sh
site/build
site/.docusaurus
.env
node_modules

View file

@ -22,6 +22,7 @@ here is a list of admin commands related to deploying .gb* files.
| Name | Description|
| --------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| Admin Notify E-mail | E-mail used to notify administrators. |
| Answer Mode | How LLM is used. Can be "direct", "document", "document-ref", "sql", "tool". |
| Avatar Logo | Bot logo URL. |
| Can Publish | Comma separeted list of mobile numbers that can publish directly. |
| Description | Description of the Bot. |
@ -51,7 +52,104 @@ here is a list of admin commands related to deploying .gb* files.
| Synchronize Database | TRUE if TABLE keyword should *MODIFY* database. Be careful in Production storage this value should be FALSE (default). |
Note that this variables are available in every .gbdialog code, automatically.
## Config in Detail
### Admin Notify E-mail
The "Admin Notify E-mail" setting specifies the email addresses that will receive notifications regarding the bots activities and administrative updates. This feature ensures that administrators are promptly informed about any critical events or issues related to the bot's operation, allowing them to take necessary actions swiftly.
### Answer Mode
In a bot factory configuration, the integration of a Language Learning Model (LLM) is pivotal in enhancing the functionality and efficiency of the bots. The LLM can be employed in various modes depending on the requirements of the bot system. These modes include "direct," "document," "document-ref," "sql," and "tool," each offering a unique method of interaction and data handling. For instance, in the "direct" mode, the LLM is used to generate responses or actions based on real-time user input, providing immediate and contextually relevant replies. This approach is particularly useful for creating conversational agents that can handle a wide range of user queries effectively.
In the "document" mode, the LLM processes and generates responses based on pre-existing documents or content. This mode is ideal for scenarios where the bot needs to reference detailed information or provide insights derived from specific documents. The LLM analyzes the content of the documents to deliver accurate and contextually appropriate information, thereby enhancing the bot's ability to offer valuable and precise responses. This approach is particularly beneficial for applications that require in-depth knowledge or detailed guidance based on existing resources.
The "document-ref" mode builds on the document-based approach by allowing the LLM to reference and extract information from multiple documents to generate comprehensive responses. This mode is advantageous when the bot needs to synthesize information from various sources to provide well-rounded answers or perform complex tasks. In contrast, the "sql" mode leverages structured query language to interact with databases, enabling the LLM to retrieve and manipulate data based on specific queries. Lastly, the "tool" mode integrates the LLM with external tools or services, allowing the bot to extend its functionality by interacting with other applications or systems. Each mode offers a distinct advantage, allowing the bot factory to tailor its approach based on the specific needs and objectives of the bot system.
### Avatar Logo
The "Avatar Logo" is the URL of the bot's logo that represents it visually. This logo is used in user interfaces and communications to create a recognizable identity for the bot, enhancing its brand presence and user engagement.
### Can Publish
The "Can Publish" setting is a comma-separated list of mobile numbers that have the authority to publish content directly. This feature ensures that only authorized individuals can contribute or update content, maintaining control over the bot's published material.
### Description
The "Description" provides a brief overview of the bot, including its purpose, functionalities, and key features. This description helps users understand what the bot does and what to expect from its interactions.
### Enable Audio Hearing
The "Enable Audio Hearing" option indicates whether the bot can process and understand audio inputs. When set to TRUE, the bot is capable of receiving and interpreting spoken language, enhancing its interaction capabilities.
### Enable Audio Talking
The "Enable Audio Talking" setting specifies if the bot can generate and deliver spoken responses. Enabling this feature allows the bot to communicate with users through audio, making interactions more dynamic and engaging.
### Enable Spelling Checker
The "Enable Spelling Checker" option determines whether the bot should automatically check and correct spelling errors in user inputs. When set to TRUE, this feature helps maintain the quality and accuracy of interactions by addressing spelling mistakes.
### Enable Worldwide Translator
The "Enable Worldwide Translator" setting allows the bot to support multiple languages and perform translation tasks. When enabled, the bot can interact with users in various languages, catering to a global audience.
### Default User Language
The "Default User Language" specifies the language that users will encounter by default when interacting with the bot. This setting ensures that the bot communicates in the preferred language of the user, enhancing their experience.
### Default Content Language
The "Default Content Language" defines the language in which content files, such as .gbkb and .gbdialog, are written. This setting ensures consistency in content creation and processing within the bot's system.
### Language Detector
The "Language Detector" feature enables the bot to automatically identify the language of user inputs. When activated, the bot can adapt its responses based on the detected language, improving communication accuracy.
### NLP Score
The "NLP Score" is a numerical value between 0 and 1 used to assess the relevance and quality of text for Natural Language Processing (NLP) neural networks. This score helps in classifying and processing text more effectively.
### Search Score
The "Search Score" is a value between 0 and 1 that influences the consideration of Full Text Search as a content retrieval method. A higher score indicates a greater emphasis on search-based content retrieval.
### Transfer To
The "Transfer To" setting is a comma-separated list of mobile numbers to which the bot will transfer chats from WhatsApp. This feature facilitates the rerouting of conversations to designated contacts or support agents.
### WhatsApp Admins
The "WhatsApp Admins" field is a comma-separated list of mobile numbers designated as administrators for the bot. These individuals have elevated access and control over the bots settings and operations.
### Feedback Improve.Message
The "Feedback Improve.Message" is a predefined message that the bot will send to users when it fails to find the requested information. This message encourages users to provide feedback or try different queries.
### Keep Text
The "Keep Text" setting is a comma-separated list of words or phrases that should remain unchanged during the bots understanding process. This feature ensures that specific terms are preserved and accurately recognized.
### Start Dialog
The "Start Dialog" specifies the name of the .docx file (without the extension) that serves as the entry point for the bots dialog. This document outlines the initial interactions and flow of the conversation.
### HelloGoodX
The "HelloGoodX" option determines whether the bot should greet users with a welcome message. When set to TRUE, the bot will initiate interactions with a friendly hello.
### XRM Key
The "XRM Key" is a string key used to authenticate and access the HubSpot API. This key is essential for integrating the bot with HubSpot services and functionalities.
### WhatsApp Group Name
The "WhatsApp Group Name" is the name of the WhatsApp group to which the bot belongs. This setting helps in identifying and organizing the bots presence within specific groups.
### WhatsApp Group Shortcuts
The "WhatsApp Group Shortcuts" are space-separated triggers in text that activate the bot within WhatsApp groups. These shortcuts enable users to engage with the bot using predefined commands or keywords.
### Bot Admin Number
The "Bot Admin Number" is the mobile number that receives debug screenshots and other notifications related to the bots operation. This contact is crucial for monitoring and troubleshooting the bot.
### WhatsApp Group ID
The "WhatsApp Group ID" is a unique identifier used to connect the bot to a specific WhatsApp group. This ID is accessible only in internal logs and is essential for group integration.
### Debug Web Automation
The "Debug Web Automation" setting indicates whether the Bot Admin Number should receive screenshots of each web automation step. Enabling this feature provides detailed visibility into web automation processes.
### .gbapp List
The ".gbapp List" is a collection of associated .gbapp packages linked to the .gbot. This list helps in managing and organizing the bots application packages.
### Group Spell
The "Group Spell" option enables the spell checker for group interactions. When activated, the bot will check and correct spelling errors in group chat messages.
### Synchronize Database
The "Synchronize Database" setting determines whether the TABLE keyword should modify the database. Setting this to TRUE allows changes to the database schema, while FALSE should be used in production environments to prevent unintended modifications.
---
Each title and description has been crafted to clarify the purpose and functionality of the respective configuration setting.
## Enviroment Variables Reference
| Name | Sample Value | Description |