@ejunior95/formai-core 🤖

The headless, framework-agnostic engine for AI-powered form generation.

This is the core library that powers the formAI ecosystem. It contains the essential logic for communicating with the formAI API, fetching AI-generated field configurations, and validating values.

Why formAI?

Stop wasting time hunting for validation regex, memorizing mask patterns, or manually building complex validation logic.

formAI lets you build forms by simply describing what you want in plain English.

• You write: "A required Brazilian phone number"
• You get:

  {
    "type": "mask-text",
    "mask": "(00) 00000-0000",
    "placeholder": "(00) 00000-0000",
    "validation": {
      "required": true,
      "regex": "^\\(\\d{2}\\)\\s\\d{5}-\\d{4}$",
      "minLength": 15,
      "maxLength": 15
    }
  }

Installation

npm install @ejunior95/formai-core

How to Use (Core Engine)

This package is "headless" and intended to be used by framework adapters (like @ejunior95/formai-react). However, you can use it directly in any JavaScript project.

import { getFieldConfig, validateValue } from '@ejunior95/formai-core';

// 1. Fetch the configuration from the AI
const config = await getFieldConfig(
  "A required email field", // The prompt
  {
    maskPatterns: { digit: '0', letter: 'a' } // Your library's mask definitions
  }
);

/*
config = {
  type: 'text',
  placeholder: '[email protected]',
  mask: null,
  validation: {
    required: true,
    regex: '^[\\w-\\.]+@([\\w-]+\\.)+[\\w-]{2,4}$',
    ...
  }
}
*/

// 2. Use the pure validator
const error = validateValue("test@", config);
// error = "Invalid format."

const success = validateValue("[email protected]", config);
// success = null

API

getFieldConfig(prompt: string, options?: FormAIOptions)

• prompt: The natural language description (e.g., "a 5-digit postal code").
• options:(Optional) An optional object.

• maskPatterns:

  • digit: The character your mask library uses for digits (e.g., '0').
  • letter: The character your mask library uses for letters (e.g., 'a').

validateValue(value: string, config: FormAIConfig)

• value: The current value of the input.
• config: The FormAIConfig object returned from getFieldConfig.

Returns null if valid, or an string with the error message if invalid.

Framework Adapters

This package is just the engine. For easy integration, use one of our official adapters:

• ⚛️ React: npm install @ejunior95/formai-react (Available Now!)
• 🅰️ Angular: Coming Soon!
• 💚 Vue: Coming Soon!