Scriptcase AI

Scriptcase AI is an assistant integrated into Scriptcase that uses artificial intelligence models to support developers in creating and maintaining their projects.

Because it is a native integration, Scriptcase AI can access project information, such as table structures, when allowed, application code, and tool resources. This access is defined by the permissions configured by the user, allowing the assistant to provide responses that are more aligned with the context of the development environment.

Scriptcase AI can be configured with the following AI providers:

• OpenAI (ChatGPT): see how to generate the ChatGPT API key.
• Google Gemini: see how to generate the Gemini API key.

Scriptcase AI Configuration

To enable the use of Scriptcase AI, you must configure the API key of the selected AI provider and define the access permissions to the project information.

To configure it, go to Tools > Scriptcase AI.

API Key

Enter the API key of the AI provider that will be used by Scriptcase AI.
This key allows Scriptcase to communicate with the selected provider to process the requests made by the user.

See the documentation corresponding to the selected provider:

• OpenAI (ChatGPT): see how to generate the ChatGPT API key.
• Google Gemini: see how to generate the Gemini API key.

Provider

Defines the AI provider that will be used by Scriptcase AI to process the requests sent by the user.

The available providers are:

• OpenAI (GPT)
• Google Gemini

Show smaller/older models

Enables the display of smaller or previous models from the selected provider.

When this option is enabled, the Chat model and Smart tasks model fields may display additional models, which can be used according to the project needs.

Chat model

Defines the AI model used in conversations with Scriptcase AI.

This model will be used to answer questions, analyze project information, and assist the user during development.

For this option, we recommend using the most up-to-date models from the selected provider, ensuring more complete and contextualized responses during conversations.

Smart tasks model

Defines the AI model used to run Scriptcase Smart tasks.

This model will be used in automatic or assisted features that perform specific tasks within the tool.

For smart tasks, we recommend using smaller models when available. In general, these models can offer lower cost and faster responses, making them suitable for simple, direct operations that do not require complex analysis.

Automatically create and translate data dictionary

Enables the automatic creation of the data dictionary and the language files used to translate field labels during project creation.

With this option enabled, Scriptcase AI can assist in generating field descriptions and translations, reducing manual work during the initial project configuration.

See the page Creating a new project for more information.

Access permissions by project

In the project list, you can configure which information Scriptcase AI will be able to access in each project.

Permissions can be enabled independently. This way, you can allow access only to the application code, only to the project connections, or enable both types of access.

Even without access permissions to the project, Scriptcase AI can still help with general questions about Scriptcase, such as using macros, events, applications, and other tool resources. It can also help with code and SQL, as long as the user manually provides the code, SQL query, or table structure they want to analyze in the chat.

These settings are available only to Scriptcase administrators.

Project

Displays the existing projects in the Scriptcase environment, allowing you to define Scriptcase AI access permissions for each project.

Code analysis

Defines whether Scriptcase AI will be able to automatically access existing code in the project’s application events, methods, and buttons.

With this option enabled, the assistant will be able to analyze the code to help with tasks such as explaining how existing snippets work, adding code to events, identifying possible errors, and suggesting syntax or logic corrections.

Note: If this option remains disabled, Scriptcase AI will still be able to help with code, as long as the content is manually provided in the chat.

Developer prompt

Allows you to enter rules, context, and standards that Scriptcase AI should consider when assisting with the project.

This field can be used to guide the responses, suggestions, and code generated by the assistant according to the project needs. For example, you can provide the system context, variable naming standards, development rules, and preferences for how responses should be presented.

Connections

Defines whether Scriptcase AI will be able to access the table structure of the connections configured in the project, such as table names, fields, and data types.

With this option enabled, the assistant will be able to generate suggestions, analyses, and SQL commands that are more aligned with the database of the enabled connections.

Note: If this option remains disabled, Scriptcase AI will still be able to help with SQL commands and database structures, as long as the information is manually entered in the chat.

Connection list

When clicking the connection icon, a modal will be displayed with the list of existing connections in the project. Enable the connections that Scriptcase AI will be able to access.

The icon color indicates the access status to the project connections.

• Gray icon - No project connection is enabled for access by Scriptcase AI.
• Yellow icon - Some project connections are enabled for access by Scriptcase AI.
• Green icon - All project connections are enabled for access by Scriptcase AI.

Using Scriptcase AI

Scriptcase AI can be used in two ways within Scriptcase:

• Assisted chat: allows you to interact directly with the AI to get support during application development.
• Smart tasks: allows you to run AI automations to speed up specific platform processes.

Assisted chat

To access the chat, click the button located in the upper-right corner of the interface.

Assisted chat allows the developer to interact directly with Scriptcase AI during application development.

The AI can use the context of the project, applications, open events, and, when allowed, the database connections and structures to provide responses that are more aligned with the development environment.

Among the features available in the chat are:

• Analyze code from open events, buttons, and methods;
• Generate code using Scriptcase macros;
• Create business rules;
• Explain how to use Scriptcase features.

Chat usage tips

• Make direct requests about the content open in the editor: ask for actions such as “fix the logic of this event” or “optimize this SQL query”. When code analysis is enabled, there is no need to copy and paste the code into the chat, as the assistant will be able to access the open content.
• Take advantage of access to table structures: when connection permission is enabled, you can request code or queries based on the project tables, such as “create a validation using the fields from the pedidos table”.
• Use AI for repetitive tasks: request the creation of validations, date and time formatting, business rules, or code snippets that can reduce manual work during development.

Chat navigation

Scriptcase AI has three main tabs to help navigate and organize the interactions performed in the chat.

• Current conversation: area used for direct interaction with the AI.
• Suggestions: displays examples of questions and commands that can be used in the chat.
• History: shows previous conversations organized in chronological order.

Smart tasks

Smart tasks use AI models to automate specific operations in Scriptcase.

Currently, the feature can be used for:

• Automatic application translation;
• Automatic data dictionary creation;
• Automatic data dictionary translation.

Because they use more specific and automated tasks, we recommend using smaller and faster models when available. In general, these models may offer lower cost and more agile responses during operation processing.