AI RAG Wrapper - Complete Documentation

Introduction

Welcome to AI RAG Wrapper, a professional Retrieval-Augmented Generation (RAG) backend built with Laravel 12. This system turns your documents into searchable knowledge for AI, enabling intelligent, context-aware interactions with your data.

License: This product requires a valid Envato Purchase Code for activation during the installation process.

What is RAG?

Retrieval-Augmented Generation (RAG) combines your private documents with AI language models. When users ask questions, the system retrieves relevant chunks from your documents and feeds them to the AI for accurate, context-aware responses.

🚀 Quick Start

5-Minute Quick Start

Step 1

Extract the source code to your server, create a blank MySQL database, and rename or copy .env.example to .env.

Step 2

Visit /install to verify your license and set up your super admin account.

Step 3

Set GEMINI_API_KEY in Settings → RAG Settings for free embeddings (recommended for getting started).

Step 4

Start the queue worker: php artisan queue:work (Required for document processing!)

Step 5

Upload your first document and click Process to begin indexing.

✅ You're ready! Visit the admin dashboard or use the API to start chatting with your documents.

Key Features

🤖 AI & RAG Engine

Smart Chunking
Intelligent text splitting with configurable chunk size (1000 chars) and overlap (100 chars) for maximum AI accuracy.
Semantic Search
Vector embeddings enable high-accuracy retrieval of document segments based on meaning, not just keywords.
Reranking Pipeline
Optional reranking stage uses specialized models to score and re-order retrieved chunks for improved precision.
Query Rewriting
Dynamically modifies user queries to better align with document language and structure.
Semantic Caching
Reduces latency and API costs by caching semantically similar previously answered questions (24-hour cache).
Streaming Responses
Real-time typewriter-style AI responses via Server-Sent Events (SSE) for better user experience.
Exponential Backoff Retry
Automatic retry logic (3 attempts, 2000ms delay) for transient AI provider failures.
Chat Rate Limiting
Configurable limit on conversation history messages to control context window size and API costs.

📄 Document Management

Multi-format Support
Upload PDF and plain text/markdown files for ingestion.
Automatic Text Extraction
PDF parsing via smalot/pdfparser with intelligent text extraction.
Background Processing
Queued job batches for scalable document ingestion without blocking the UI.
Real-time Progress
Track document processing status with percentage completion indicators.
Custom Instructions
Per-document system prompts and instructions for tailored AI behavior.
UUID Identification
Secure, non-sequential document identifiers for enhanced security.
Cascade Deletion
Automatic cleanup of associated knowledge chunks on document deletion.

⚙️ Technology Stack

Laravel 12
Latest LTS framework with modern PHP 8.3+ features.
Filament v5
Enterprise-grade admin panel with beautiful UI components.
Livewire 4
Full-stack framework for dynamic interfaces without JavaScript.
Flux UI
Modern, reactive UI components with Tailwind CSS v4 styling.
Pest 4
Expressive testing framework ensuring 100% feature reliability.
Tailwind CSS 4
Utility-first CSS for premium design aesthetics.

System Requirements

Component	Requirement
PHP Version	8.3+
Database	MySQL 8.0+ or PostgreSQL
Required Extensions	BCMath, Ctype, Fileinfo, JSON, Mbstring, OpenSSL, PDO, Tokenizer, XML
Memory Limit	256MB+ (Required for PDF parsing)
SSE Requirements	`mod_security` disabled or exception made; `fastcgi_buffering off` for Nginx

Installation Guide

Standard Installation

Upload: Zip the main_files folder and upload the mail_files.zip to your server.
Configure Environment: As for CodeCanyon, the .env file will not be present. Create a .env file by doing ONE of the following:
- Option A (Recommended): Rename .env.example to .env
- Option B: Copy .env.example and save the copy as .env
- Option C: Create a new blank file named .env and paste the contents of .env.example into it
💡 Tip: The .env file stores your database credentials, API keys, and app settings. Never share this file publicly.
Run Wizard: Navigate to your-domain.com/install.
License: Enter your Purchase Code. The system will verify this with the Envato API.
Database: Enter your credentials and let the wizard migrate the tables.
Super Admin: Create your first administrator account.

What Happens During Installation

The installation wizard automatically configures the complete Role-Based Access Control (RBAC) system using Filament Shield. When you create your first administrator account, the system performs the following actions behind the scenes:

✅ Shield Installation

Runs shield:install to create the Roles resource and initial permission structure in Filament.

🔐 Permission Generation

Executes shield:generate --all to automatically generate permissions for all Filament resources (Resources, Pages, and Widgets).

👑 Super Admin Setup

Runs shield:super-admin to assign the super_admin role to your first user, granting full access to all features.

🛡️ Default Role Creation

Creates a default user role with basic permissions for regular users.

🎯 Complete RBAC Setup: The installer handles the entire Role-Based Access Control configuration automatically. After installation, you can manage roles and permissions through the Filament admin panel at your-domain.com/admin/settings/roles.

License Verification Behavior

The installer verifies your purchase code with the Envato API. To ensure legitimate users aren't blocked during API outages while still protecting against invalid licenses, the system implements graceful degradation with distinct behavior for installation vs. runtime:

Scenario	Phase	Behavior	Why?
API Down (503/502)	Installation	⚠️ Allow + Warning	Records purchase code, marks as unverified. Auto-retries later. Prevents blocking legitimate users during outages.
API Down (503/502)	Runtime	✅ Allow (Cached)	If previously verified, use cached license. Does not lock users out during daily use.
Rate Limited (429)	Installation	⚠️ Allow + Warning	Records purchase code, waits for rate limit to reset. Auto-retries after cooldown.
Rate Limited (429)	Runtime	✅ Allow (Cached)	Relies on local cache until rate limit resets. No user disruption.
Connection Error	Installation	⚠️ Allow + Warning	Network issues shouldn't block installation. Purchase code recorded for later verification.
Invalid Code (404)	Installation	❌ Block	API explicitly returns "Invalid". User must check and re-enter code.
Invalid Code (404)	Runtime	⚠️ Log Warning	Previously valid license now invalid? Log and alert admin, but don't abruptly lock out.
Wrong Item ID	Installation	❌ Block	Purchase code is for a different product. User must verify correct item.
Valid License	Runtime	✅ Allow	License verified and cached. Full access granted.

🔄 Automatic Re-verification: If the API is unavailable during installation, the system records your purchase code and automatically attempts verification later (once per hour). The application functions normally in the meantime.

Key Design Principles

✅ Never Block During API Outages

Legitimate users should not be locked out when Envato API experiences downtime. Cache and grace periods protect users.

❌ Always Block Invalid Licenses

When the API explicitly returns "Invalid", block immediately to prevent unauthorized use.

⚠️ Throttle Runtime Checks

Runtime verification runs at most once per hour to avoid rate limits and minimize API dependency.

Detailed cPanel Installation Guide

The installation of this script is super easy. If you can install normal php based scripts on your server, you can install our script. We will tell you how to install this script step by step on a server.

If you extract the zip file, you will see that there are total two folders.

The folder names are: documentation and main_files. The "main_files" folder contains the complete Laravel source code.

Creating Database using Cpanel

Login to your hosting's cpanel using the login information. You will see an interface like this:
You will get a search section in the cpanel. Search there by writing the word "database". You will get some options. Click on "MySQL® Database Wizard" option.
Then in Step 1, you will have to create a database. In the text field, give your database name and click on next step.
Then in Step 2, you will have to create a database user. Give your username and passwords here. Then click on "Create User" button.
Now a permission page will come. Just check the tick mark on "ALL PRIVILEGES" and you will see all the options will be automatically checked.
Click on the "Next Step" button in the bottom.
Done! Your database and user are now ready. The application installer will automatically create all necessary tables and initial data during the setup process in the next section.

Uploading Files using Cpanel

In the search field of your cpanel, write the word "file". You will get a "File Manager" option. Click on that option.
Search for the "public_html" and go into there.
Now go to the folders that you got after extracting the main zip and compress the folder "main_files" using any software like winrar.
The zip file will be like this.
In the cpanel's "public_html", upload this zip file.
You will get an interface like this.
Drag and drop the zip file here or just click on the "Select File" and select the file from your computer. You will see a progress of uploading file. Wait until it is finished.
After finishing the upload, you will see a green bar. It indicates that upload is successful.
Now close the upload window. In the public_html section, press on "Go" button to refresh the page. Then you will see the zip file in the right side.
Do right click on that file and select "Extract" option.
Press on "Extract Files" button.
It will take only a few minutes. After finish, you will get a window like this.
Just close this window. Then double click on the "main_files" folder and go into there.
Now click on the "Select All" from top section.
Click on the "Move" option.
Delete the "main_files" from the box. Only keep this "/public_html/". Then click on "Move Files".
Now, click on "Up One Level".

Renaming .env.example to .env

Follow these steps to rename the .env.example file to .env:

Right click on the .env.example file and click on Rename

Remove .example and keep only .env

Click Rename File

Your task is done and you can use our script without any problem. :)

💡 Next Step: Navigate to your-domain.com or your-domain.com/install to access the installation wizard and complete the setup.

⚠️ Crucial: Don't forget to follow the Background Processing & Scheduled Tasks (Crucial) section below. Without this setup, document processing will stay at 0% and your documents won't be ingested!

Uploading Files using FTP

Download any FTP client on your computer like "FileZilla". Search on Google by "FileZilla" or go to their website to download: https://filezilla-project.org/
Use the first option to download
After download, install it and open. The interface will be like this:
You can collect your FTP information contacting your hosting provider. Or, going to cpanel, you can create your FTP. In cpanel, search for "FTP".
Put ftp username, password in the form. Become sure, you delete the marked box data in the following image.
Click on "Create FTP Account" button.
Your FTP information will be like this:
FTP Host: {your hosting provider will give you host name}
FTP Port: 21
FTP Username: ftp@domain.com
FTP Password: {that you gave while creating the FTP}
Now in filezilla put those information and click on "Quickconnect" button.
If connection is successful, you will get an interface like this:
Now in the left side box, go to the location where you have extracted the script's file. And in the right side, go to public_html.
Then right click on the "main_files" from left side box and click on "Upload". All the files then will be uploaded into the "public_html" of the right side.

📝 Note: Don't forget to rename or copy .env.example to .env after uploading. See the cPanel Settings section in the Installation Guide for detailed instructions on how to rename the .env file via cPanel.

Background Processing & Scheduled Tasks (Crucial)

The RAG Engine uses background jobs to process documents. If this is not set up, document ingestion will stay at 0%.

Recommended: Laravel Scheduler (Single Cron Job)

Add this single Cron Job to run every minute. This handles queue processing:

* * * * * /usr/local/bin/php /home/username/public_html/artisan schedule:run

The scheduler automatically manages:

Queue Processing - Runs every minute to process document jobs

Alternative: Direct Queue Worker

If you prefer not to use the scheduler, run the queue worker directly:

* * * * * /usr/local/bin/php /home/username/public_html/artisan queue:work --stop-when-empty

VPS (Supervisor for Continuous Processing)

For VPS environments, use Supervisor for continuous queue processing:

php artisan queue:work --tries=3

Shared Hosting (Cron Job)

Add a Cron Job in cPanel to run every minute:

⚠️ Important: Replace the paths with your actual server paths:

PHP Path: Replace /usr/local/bin/php with your actual PHP path (e.g., /usr/bin/php8.3). Find it using whereis php
Project Path: Replace /home/username/public_html/ with your actual project path. Most users upload to public_html/, but if you uploaded to a subfolder, use /home/username/public_html/your-project/artisan

Common Paths by Hosting:

cPanel (public_html): /home/username/public_html/artisan
cPanel (subfolder): /home/username/public_html/myproject/artisan
Plesk (httpdocs): /var/www/vhosts/username/httpdocs/artisan
DirectAdmin: /home/username/domains/your-domain.com/public_html/artisan

Installation Wizard Steps

Go to your Domain- your-domain.com and you will see the installation wizard in the your-domain.com/install url. The installation wizard guides you through the setup process in 5 simple steps. Here's what to expect at each stage:

Step 1: Preflight Checks

The system automatically verifies that your server meets all requirements:

PHP Version (>= 8.3) - Ensures compatibility with modern PHP features
Required Extensions - Checks for necessary PHP extensions (BCMath, Ctype, Fileinfo, JSON, Mbstring, OpenSSL, PDO, Tokenizer, XML)
Directory Permissions - Verifies write access to storage and cache directories
.env File Writable - Confirms the environment file can be modified

All checks must pass before you can proceed. Click Continue to move to the next step.

Step 2: License Verification

Enter your Envato/CodeCanyon purchase details to verify your license:

Purchase Code - Enter the license code you received from CodeCanyon (format: XXXX-XXXX-XXXX-XXXX)
The system will verify this code with the Envato API to ensure it's valid and matches this product

If the Envato API is temporarily unavailable, the system will allow installation with a warning and verify the license later.

Click Verify License to proceed.

Step 3: Database Configuration

Provide your database connection details:

Database Host - Typically 127.0.0.1 or localhost
Database Port - Default MySQL port is 3306
Database Name - The name of the database you created for this application
Database Username - Your database username (e.g., root)
Database Password - Your database password (leave empty if no password)

The wizard will automatically run database migrations to create all required tables after you click Configure Database.

Step 4: Application Configuration

Configure your basic application settings:

Application Name - The name that will appear in the browser title and throughout the app (e.g., "My App")
Application URL - The full URL where your application is hosted (e.g., https://your-domain.com)

Click Save Configuration to proceed to the final step.

Step 5: Create Super Admin

Create the initial administrator account:

Full Name - Your name or the admin user's name
Email Address - The admin email (used for login)
Password - Choose a strong password (minimum 8 characters recommended)
Confirm Password - Re-enter the same password

This user will have the super_admin role with full access to all features. Click Complete Installation to finish setup.

Installation Complete!

After completing all 5 steps, you'll be redirected to the admin dashboard. You can now:

Configure AI providers in Settings → RAG Settings
Upload your first document
Start using the RAG chat features

Next Step: Don't forget to start the queue worker by running php artisan queue:work to enable document processing.

🔧 Troubleshooting

Common issues and their solutions. Check this section before opening a support ticket.

Common Issues

Issue Ingestion stuck at 0%

Fix: Ensure php artisan queue:work is running. Document processing requires a queue worker to be active.

Check if queue worker is running: ps aux | grep queue:work
Restart the queue: php artisan queue:restart then start a new worker
Check failed jobs: php artisan queue:failed

Issue 404 errors on API routes

Fix: Ensure mod_rewrite is enabled and .htaccess is in the root directory.

Apache: Enable mod_rewrite (a2enmod rewrite on Ubuntu)
Ensure .htaccess file exists in your public root
Check AllowOverride All in Apache config
Nginx: Ensure proper rewrite rules are configured

Issue PDF parsing fails

Fix: Increase memory_limit to 512MB in php.ini.

Edit php.ini: memory_limit = 512M
Or add to .htaccess: php_value memory_limit 512M
Or add to .user.ini: memory_limit = 512M
Restart your web server after changes

Issue "Class not found" errors

Fix: Run composer install --optimize-autoloader to regenerate autoload files.

Issue Permission denied errors

Fix: Set correct permissions on storage and cache directories.

chmod -R 775 storage bootstrap/cache
chown -R www-data:www-data storage bootstrap/cache

Issue Email verification not sending

Fix: Configure mail settings in .env file.

Set MAIL_MAILER (smtp, mailgun, ses, etc.)
Configure SMTP credentials if using SMTP
Test with: php artisan mail:test

Issue AI responses are slow

Tip: Enable semantic caching to reduce API calls and improve response times for repeated queries.

Go to Settings → General Settings
Toggle Enable Semantic Cache to ON
Consider using Groq or OpenRouter for faster inference

Getting Help

If you've checked the troubleshooting steps above and still need assistance:

Check the logs: storage/logs/laravel.log
Enable debug mode: Set APP_DEBUG=true temporarily in .env
Contact support: Include your PHP version, error logs, and steps to reproduce

Admin Dashboard (Filament)

A enterprise-grade administrative suite built with Filament v5 for total system control.

Navigation Structure

Section	Resources
Dashboard	API Health Stats, Latest Documents, Token Counter, Job Monitor
Users	User Management, Role Management
API Tokens	Manage Sanctum Personal Access Tokens
Documents	Document Manager, Knowledge Chunks
Settings	General Settings, RAG Settings

Dashboard Widgets

API Health Check: Verifies connectivity to configured AI providers and database services with tier detection (Free/Premium).
API Route Explorer: Interactive search and filtering of all available API endpoints.
Job Monitor: Visual breakdown of background job success, failures, and execution times.
Token Counter: Widget displaying token usage statistics.
Latest Documents: Widget showing recently uploaded and processed documents with status indicators.
API Health Stats: Real-time statistics on API health check results (healthy, rate-limited, errors).

Management Resources

Complete reference for all administrative resources with their inputs, tables, and available actions.

👥 Users Resource

Complete CRUD interface for managing system users with role assignment and permission management.

Form Inputs (Create/Edit)

Field	Type	Validation	Description
`name`	Text Input	Required	User's full name
`email`	Email Input	Required, Unique, Valid Email	User's email address
`password`	Password Input	Required (Create only), Hashed	User's password (auto-hashed)
`roles`	Multi-Select	Optional	Assign one or more roles to user

Table Columns

Column	Sortable	Searchable	Description
Name	✓	✓	User's full name
Email	✓	✓	Email address
Roles	-	-	Assigned roles (badge separator)
Created/Updated At	✓	-	Timestamps (toggleable)

Available Actions

View: Slide-over modal with full user details
Edit: Slide-over modal for quick edits
Delete: Remove user with confirmation
Bulk Delete: Delete multiple users at once

API Endpoints

GET/api/users - List all users (admin only)
GET/api/users/{user} - Get user details
POST/api/users - Create user
PUT/api/users/{user} - Update user
DELETE/api/users/{user} - Delete user
GET/api/me - Get current user profile

🔐 Roles Resource (Shield)

Advanced RBAC system with granular permission management via Filament Shield.

Form Inputs (Create/Edit)

Field	Type	Validation	Description
`name`	Text Input	Required, Unique, Max 255	Role name (auto-formatted)
`guard_name`	Select	Required	Auth guard (web/sanctum)
`team`	Select	Optional	Team association (if tenancy enabled)
`Select All`	Toggle	-	Select/deselect all permissions
`Permissions`	Checkbox Grid	Optional	Granular permissions by resource

Table Columns

Column	Sortable	Searchable	Description
Name	-	✓	Role name (headline formatted)
Guard Name	-	-	Auth guard badge
Team	-	✓	Team association (if visible)
Permissions Count	-	-	Number of assigned permissions
Updated At	✓	-	Last modification timestamp

Available Permissions

view - View resource
create - Create new records
update - Edit existing records
delete - Soft delete records
restore - Restore deleted records
force-delete - Permanently delete records

API Endpoints

GET/api/roles - List all roles
GET/api/roles/{role} - Get role details
POST/api/roles - Create role
PUT/api/roles/{role} - Update role
DELETE/api/roles/{role} - Delete role
POST/api/roles/{role}/permissions - Sync permissions
GET/api/permissions - List all permissions

🔑 API Tokens Resource

Manage Laravel Sanctum Personal Access Tokens with full integration into the Shield permission system.

Resource Features

Revoke Tokens: Instantly invalidate any active API token.
Ownership Logic: Users can only view and revoke their own tokens (managed via Policy).
Admin Oversight: Super admins can view and manage all system tokens.
Shield Integration: Granular permissions controlled via the Roles resource.

Table Columns

Column	Sortable	Searchable	Description
Name	✓	✓	Token descriptive name (e.g., "Mobile App")
User	-	✓	The token owner (email)
Last Used At	✓	-	Human-readable relative time (e.g., "2 hours ago")
Created At	✓	-	Formatted creation date

Available Actions

Revoke Token: Individual delete action renamed to "Revoke Token" with danger confirmation.
Bulk Revoke: Revoke multiple tokens at once via bulk actions.

Authorization

Access is strictly controlled via PersonalAccessTokenPolicy and Shield permissions:

ViewAny:PersonalAccessToken - Required to list tokens.
Delete:PersonalAccessToken - Required to revoke tokens.

📁 Documents Resource

Manage uploaded files, view processed text chunks, and track ingestion progress in real-time.

Form Inputs (Create/Edit)

Field	Type	Validation	Description
`uuid`	Text Input	Disabled, Auto-generated	Document ID (auto UUID)
`name`	Text Input	Required	Document title/name
`system_prompt`	Textarea	Optional	AI behavior definition with template picker
`instructions`	Textarea	Optional	AI tool description for document usage
`path`	File Upload	Required, PDF/TXT/MD	File upload (local disk, knowledge directory)
`status`	Select	Required	Status: pending, processing, completed, failed
`job_batch_id`	Text Input	Disabled	Background job batch identifier
`error`	Textarea	Disabled	Error message (if failed)
`user_id`	Hidden	Auto-filled	Current user ID

Table Columns

Column	Sortable	Searchable	Description
Document ID (UUID)	-	✓	Copyable UUID with icon
Name	-	✓	Document filename
Status	-	-	Badge with color + icon (animated for processing)
Progress	-	-	Real-time percentage (5s polling)
User	✓	-	Document owner
Created At	✓	-	Upload timestamp (toggleable)

Available Actions

Process: Trigger document ingestion (hidden if processing/completed)
Chat: Document-specific chat with slide-over modal
Download: Download original file
View: Full document details with chunks relation manager
Edit: Edit document metadata
Process All: Bulk process all pending/failed documents
Bulk Delete: Delete multiple documents

Relation Managers

Chunks: View all knowledge chunks extracted from document

API Endpoints

GET/api/v1/documents - List all documents
GET/api/v1/documents/{document} - Get document details
DELETE/api/v1/documents/{document} - Delete document
GET/api/v1/documents/batch/{batchId} - Get batch processing status

Resource Response Format

{
    "id": "uuid-string",
    "title": "document-name.pdf",
    "file_type": "pdf",
    "status": "completed",
    "progress": "100%",
    "created_at": "2026-01-01T00:00:00Z",
    "human_readable_size": "1.5 MB"
}

🧩 Knowledge Chunks Resource

Browse and search individual knowledge chunks extracted from documents with search & replace functionality.

Form Inputs (Edit)

Field	Type	Validation	Description
`content`	Textarea	Required	Chunk text content (full column span)
`source`	Text Input	Disabled	Source document reference
`status`	Select	Disabled	Status: syncing, synced, failed

Table Columns

Column	Sortable	Searchable	Description
Content	-	✓	Limited to 100 chars, expandable
Status	✓	-	Badge with color + icon
Source	-	✓	Source document (toggleable)
Created At	✓	-	Creation timestamp (toggleable)

Filters

Status Filter: Filter by syncing, synced, or failed status

Available Actions

Edit: Slide-over modal for chunk editing
Delete: Remove individual chunk
Search & Replace: Bulk search and replace across all chunks
Bulk Delete: Delete multiple chunks

Search & Replace Input

Field	Type	Validation	Description
`search`	Text Input	Required	Text to find
`replace`	Text Input	Required	Replacement text

💬 RAG Chat Testing

You can test the chat functionality with your uploaded documents by visiting the following URL:

Test URL: your-domain.com/markdownrag

Settings Panels

Centralized configuration interfaces for system-wide settings.

⚙️ RAG Settings

Configure AI providers, API keys, and RAG pipeline behavior.

Configuration Sections & Inputs

Website Information

Field	Type	Default	Description
`APP_NAME`	Text Input	Laravel	Application name

Database Configuration (2 columns)

Field	Type	Default	Description
`DB_CONNECTION`	Text Input	mysql	Database driver
`DB_HOST`	Text Input	127.0.0.1	Database host
`DB_PORT`	Numeric Input	3306	Database port
`DB_DATABASE`	Text Input	laravel	Database name
`DB_USERNAME`	Text Input	root	Database username
`DB_PASSWORD`	Password (Revealable)	-	Database password

General Configuration (2 columns)

Field	Type	Description
`AI_DEFAULT_PROVIDER`	Select (Native: false)	Default LLM provider for chat/completion
`AI_EMBEDDING_PROVIDER`	Select (Native: false)	Provider for vector embeddings

AI Provider Keys (2 columns)

Field	Type	Provider
`OPENAI_API_KEY`	Password (Revealable)	OpenAI (GPT-4, GPT-3.5, embeddings)
`ANTHROPIC_API_KEY`	Password (Revealable)	Anthropic (Claude)
`GEMINI_API_KEY`	Password (Revealable)	Google Gemini (Free tier available)
`GROQ_API_KEY`	Password (Revealable)	Groq (Fast inference)
`COHERE_API_KEY`	Password (Revealable)	Cohere (Embeddings)
`DEEPSEEK_API_KEY`	Password (Revealable)	Deepseek
`MISTRAL_API_KEY`	Password (Revealable)	Mistral AI
`OPENROUTER_API_KEY`	Password (Revealable)	OpenRouter (Multi-model)
`VOYAGEAI_API_KEY`	Password (Revealable)	VoyageAI (Embeddings/Reranking)
`XAI_API_KEY`	Password (Revealable)	xAI (Grok)
`JINA_API_KEY`	Password (Revealable)	Jina AI (Embeddings)
`ELEVENLABS_API_KEY`	Password (Revealable)	ElevenLabs (Voice/Speech)
`OLLAMA_API_KEY`	Text Input	Ollama (Local/Self-hosted)

RAG Pipeline Settings

Field	Type	Default	Description
`MARKDOWN_RERANKING`	Toggle	false	Enable reranking model for better precision
`MARKDOWN_QUERY_REWRITE`	Toggle	false	Rewrite queries to match document content
`MARKDOWN_CHAT_RATE_LIMIT`	Numeric Input	1	Max requests per minute for chat
`MARKDOWN_EMBEDDING_BATCH_SIZE`	Numeric Input	50	Documents per embedding request
`MARKDOWN_AI_RETRY_MAX_ATTEMPTS`	Numeric Input	3	Max retries for AI service calls

🔧 General Settings

Application-wide configuration for AI behavior, caching, and API rate limiting.

AI Configuration Section

Field	Type	Default	Description
`enable_semantic_cache`	Toggle	false	Cache identical query responses
`system_prompt`	Textarea (4 rows)	null	Global AI system prompt with template picker
`tool_description`	Textarea (4 rows)	null	AI tool description with template picker
`enable_stream_effect`	Toggle	true	Enable streaming animation for responses

API Rate Limiting Section

Field	Type	Default	Description
`enable_rate_limiting`	Toggle	false	Enable API rate limiting
`api_burst_limit`	Numeric Input	60	Max requests per minute
`api_sustained_limit`	Numeric Input	1000	Max requests over sustained period
`api_sustained_period`	Select	day	Period: hour or day

Template Pickers

Prompt Templates: Modal with pre-built system prompt templates
Tool Templates: Modal with pre-built tool description templates

API Reference (v1)

Complete API documentation with all endpoints, inputs, and response formats.

💡 Pro Tip: Postman Collection Included

We have included a pre-configured Postman Collection in the tests/postman folder. This collection includes:

Pre-request Scripts: Automatically handle authentication and token management
Automated Tests: Poll the batch status until 100% completion
Environment Variables: Pre-configured for easy setup
Newman Integration: Run automated tests via terminal with npm run postman:test

Postman Ready Newman Tests Auto-Auth

🧪 Testing with Postman

We have included a pre-configured Postman collection and environment located in the Extras/Postman/ folder of your download.

Import both files into Postman.
The collection is pre-set to use the postman@test.com account created during installation.
Ensure your {{base_url}} environment variable matches your local installation path.

Authentication Required

All API endpoints require authentication via Laravel Sanctum personal access tokens. Include the token in the Authorization header:

Authorization: Bearer YOUR_API_TOKEN

Authentication & Users API

Method	Endpoint	Description	Permissions
POST	`/api/login`	User authentication	Public
POST	`/api/users`	User registration	Public
GET	`/api/me`	Get current user profile	Auth
GET	`/api/users`	List all users	Admin (viewAny)
GET	`/api/users/{user}`	Get user details	Admin or Self (view)
PUT	`/api/users/{user}`	Update user	Admin or Self (update)
PATCH	`/api/users/{user}`	Update user (partial)	Admin or Self (update)
DELETE	`/api/users/{user}`	Delete user	Admin (delete)

User Roles Management

Method	Endpoint	Description	Permissions
GET	`/api/users/{user}/roles`	List user roles	Admin or Self (update)
POST	`/api/users/{user}/roles`	Assign role to user	Admin or Self (update)
DELETE	`/api/users/{user}/roles/{role}`	Detach role from user	Admin or Self (update)

Roles & Permissions API

Method	Endpoint	Description	Permissions
GET	`/api/roles`	List all roles	Admin (viewAny)
GET	`/api/roles/{role}`	Get role details	Admin (view)
POST	`/api/roles`	Create new role	Super Admin (create)
PUT	`/api/roles/{role}`	Update role	Admin (update)
PATCH	`/api/roles/{role}`	Update role (partial)	Admin (update)
DELETE	`/api/roles/{role}`	Delete role	Admin (delete)
POST	`/api/roles/{role}/permissions`	Sync role permissions	Super Admin (create)
GET	`/api/permissions`	List all permissions	Admin

Documents & RAG API (v1)

Method	Endpoint	Description	Permissions
GET	`/api/v1/documents`	List all documents	Auth (viewAny)
GET	`/api/v1/documents/{document}`	Get document details	Auth (view)
DELETE	`/api/v1/documents/{document}`	Delete document	Auth (delete)
GET	`/api/v1/documents/batch/{batchId}`	Get batch status	Auth (owner or admin)
POST	`/api/v1/ingest`	Trigger document processing	Auth (create)
POST	`/api/v1/chat/ask`	RAG chat with SSE streaming	Auth + Throttle (viewAny)

Chat Request Input

Field	Type	Required	Description
`query`	String	✓	User's question or message
`document_id`	String (UUID)	Optional	Specific document to query

SSE Stream Response Format

data: {"text": "AI response chunk"}

data: {"text": "continuing response..."}

data: {"text": "final part of response"}

Error Response

data: {"error": "Error message"}

Token Management API (v1)

Method	Endpoint	Description
GET	`/api/v1/tokens`	List personal access tokens
POST	`/api/v1/tokens`	Create new access token
DELETE	`/api/v1/tokens/{token}`	Revoke access token

⚠️ Important: When creating a new token via POST /api/v1/tokens, the API returns the plain_text_token. This is shown only once. Ensure you save it securely; if lost, the token must be revoked and recreated.

Rate Limiting

The /api/v1/chat/ask endpoint uses the rag-api throttle middleware with configurable limits:

Burst Limit: 60 requests per minute (default)
Sustained Limit: 1000 requests per day (default)

Configure in Settings > General Settings > API Rate Limiting.

Supported AI Providers

📊 AI Provider Model Comparison

Help choosing the right model for your use case:

Purpose	Recommended Model	Provider	Benefit
Embeddings	`text-embedding-004`	Gemini	Free and high accuracy ⭐
Fast Chat	`qwen/qwen-3.5-9b`	OpenRouter	Incredible speed/cost ratio
Complex Logic	`openai/gpt-4o`	OpenAI	Best for difficult document reasoning
Budget Chat	`deepseek-chat`	Deepseek	Very low cost, good quality
Ultra Fast	`llama-3.1-70b`	Groq	Fastest inference available
Long Context	`claude-3-5-sonnet`	Anthropic	200K context window

Free Tier: Gemini Best Value: OpenRouter Fastest: Groq

LLM Providers (Chat/Completion)

Provider	Models	Config Field
OpenAI	GPT-4, GPT-3.5-turbo	`OPENAI_API_KEY`
Anthropic	Claude 3, Claude 2	`ANTHROPIC_API_KEY`
Google Gemini	Gemini Pro, Gemini Ultra	`GEMINI_API_KEY`
Groq	Llama, Mixtral (Fast)	`GROQ_API_KEY`
Cohere	Command, Command-R	`COHERE_API_KEY`
Deepseek	Deepseek Chat	`DEEPSEEK_API_KEY`
Mistral	Mistral Large, Medium	`MISTRAL_API_KEY`
OpenRouter	Multi-model access	`OPENROUTER_API_KEY`
xAI	Grok-1, Grok-2	`XAI_API_KEY`
Ollama	Local models	`OLLAMA_API_KEY`

Embedding Providers

Provider	Models	Tier
Gemini ⭐	text-embedding-004	Free tier available
OpenAI	text-embedding-3-small/large	Paid
Ollama	nomic-embed-text, mxbai-embed	Free (Self-hosted)
Jina AI	jina-embeddings-v2	Free tier
Cohere	embed-english-v3.0	Paid
Mistral	mistral-embed	Paid
VoyageAI	voyage-2, voyage-3	Paid

Recommended Setup: Use Gemini for embeddings (free) and OpenRouter for chat (qwen/qwen3.5-9b).

Specialized Providers

Provider	Purpose	Config Field
VoyageAI	Embeddings & Reranking	`VOYAGEAI_API_KEY`
Jina AI	Embeddings	`JINA_API_KEY`
ElevenLabs	Voice/Speech	`ELEVENLABS_API_KEY`

Note: OpenRouter does NOT support embeddings. Use Gemini, Jina, or OpenAI for embeddings instead.

🔑 How to Configure AI Provider APIs

Step-by-Step Guide to Adding API Keys

Follow these steps to configure your AI providers through the Filament admin panel:

Step 1

Navigate to Settings

Log in to your admin panel and click on Settings in the left sidebar, then select RAG Settings from the dropdown menu.

Step 2

Locate AI Provider Section

Scroll down to the AI Providers section. You'll see configuration fields for each supported provider (OpenAI, Gemini, Anthropic, Groq, etc.).

Step 3

Enter API Keys

For each provider you want to use:

Find the provider name (e.g., "OpenAI API Key", "Gemini API Key")
Click on the input field and paste your API key
The field is masked for security (shown as dots)
Use the reveal icon (eye icon) to verify your entry if needed

Step 4

Set Default Providers

In the General Configuration section:

Default AI Provider: Select your preferred LLM provider from the dropdown (e.g., OpenAI, Gemini, Anthropic)
Default Embedding Provider: Select your embedding provider (Gemini recommended for free tier)
Default Model: Choose the specific model to use (e.g., gpt-4, gemini-pro)

Step 5

Optional: Enable Advanced Features

Toggle these features on/off as needed:

Enable Reranking: Improves search precision (requires VoyageAI or compatible provider)
Enable Query Rewriting: Optimizes queries for better results
Enable Semantic Caching: Reduces API costs by caching similar queries

Step 6

Save Configuration

Click the Save button at the bottom of the page. The system will:

Validate your API keys with the respective providers
Clear the configuration cache automatically
Display success/error messages for each provider

Step 7

Verify API Health

After saving, navigate to API Health Check page to verify:

All configured providers show as "Healthy"
API keys are valid and working
Response times are acceptable
Tier information (Free/Premium) is displayed

💡 Pro Tips:

Getting Started: Use Gemini for both embeddings and chat (free tier available)
Best Value: Gemini embeddings + OpenRouter chat (qwen models)
Maximum Privacy: Use Ollama for local, self-hosted models
API Key Security: Never share or commit API keys to version control
Multiple Providers: You can configure multiple providers simultaneously and switch between them

⚠️ Troubleshooting:

Invalid API Key error: Double-check your key and ensure it's active in your provider's dashboard
Provider not working: Check API Health page for detailed error messages
Changes not saving: Ensure your .env file has proper write permissions
Rate limit errors: Check your provider's usage limits and upgrade plan if needed

RAG Engine Configuration

Pipeline Flow

┌─────────────────────┐
│  User Query         │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Semantic Cache     │ ← Check for similar cached queries
└──────────┬──────────┘
           │ (miss)
           ▼
┌─────────────────────┐
│  Query Rewriting    │ ← Optional: Rewrite for better match
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Vector Search      │ ← Retrieve relevant chunks
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Reranking          │ ← Optional: Re-order by relevance
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  LLM Generation     │ ← Generate response with context
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Stream Response    │ ← SSE typewriter effect
└─────────────────────┘

Configuration Options

Setting	Env Variable	Default	Description
Chunk Size	-	1000 chars	Text splitter chunk size
Chunk Overlap	-	100 chars	Overlap between chunks
Embedding Batch Size	`MARKDOWN_EMBEDDING_BATCH_SIZE`	50	Chunks per embedding request
Chat Rate Limit	`MARKDOWN_CHAT_RATE_LIMIT`	1	Previous messages in context
Retry Max Attempts	`MARKDOWN_AI_RETRY_MAX_ATTEMPTS`	3	AI provider retry count
Retry Delay	-	2000ms	Exponential backoff delay
Cache TTL	-	24 hours	Semantic cache expiration

Inputs & Forms Reference

Complete list of all input types used across the application.

Input Types

Type	Usage	Features
Text Input	Names, titles, identifiers	Validation, required, unique, max length
Email Input	User emails	Email validation, unique check
Password Input	Passwords, API keys	Masked, revealable, hashed on save
Numeric Input	Ports, limits, counts	Number-only validation, min/max
Textarea	Prompts, descriptions	Multiple rows, nullable, template pickers
Select (Dropdown)	Status, providers, roles	Native/non-native, searchable, multi-select
Toggle	Boolean flags	On/off, default values
File Upload	Document uploads	File type validation, disk selection, directory
Date/Time Picker	Timestamps	Calendar picker, timezone support
Multi-Select	Roles, permissions	Multiple selection, preload options
Checkbox Grid	Permissions matrix	Bulk select, grouped by resource
Hidden Input	Auto-filled values	Default values, user_id injection

Special Input Features

Template Pickers (Modal)

Used in system prompts and tool descriptions. Opens a modal with pre-built templates:

Prompt Templates: Legal expert, Medical advisor, Technical support, etc.
Tool Templates: Document retrieval, FAQ bot, Knowledge base search

Copyable Fields

UUIDs and identifiers with one-click copy functionality:

Document UUID (with copy icon and confirmation message)

Disabled/Read-Only Fields

Auto-populated fields that cannot be manually edited:

UUID (auto-generated)
Job Batch ID (auto-assigned)
Error messages (auto-filled on failure)
Source (auto-linked to document)
Status (auto-updated by jobs)

Tables & Filters Reference

All data tables with their columns, sorting, searching, and filtering capabilities.

Table Features

Polling: Documents table refreshes every 5 seconds for real-time progress
Sortable Columns: Click column headers to sort ascending/descending
Searchable: Search icon enables text filtering on marked columns
Toggleable: Click table settings to show/hide optional columns
Record Actions: Per-row action buttons (view, edit, delete, custom)
Bulk Actions: Select multiple records for batch operations
Header Actions: Table-level actions (Process All, Search & Replace)

Documents Table

Column	Type	Sortable	Searchable	Toggleable	Special Features
UUID	Text	-	✓	-	Copyable, copy message
Name	Text	-	✓	-	-
Status	Badge	-	-	-	Color-coded, animated icon, error tooltip
Progress	Text	-	-	-	Real-time via batch lookup
User	Relation	✓	-	-	Displays owner name
Created At	DateTime	✓	-	✓	Hidden by default

Users Table

Column	Type	Sortable	Searchable	Toggleable
Name	Text	-	✓	-
Email	Text	-	✓	-
Roles	Badge List	-	-	-
Created At	DateTime	✓	-	✓
Updated At	DateTime	✓	-	✓

Roles Table

Column	Type	Sortable	Searchable	Special Features
Name	Text	-	✓	Headline formatted
Guard Name	Badge	-	-	Color: warning
Team	Badge	-	✓	Visible if tenancy enabled
Permissions Count	Badge	-	-	Counts relationship
Updated At	DateTime	✓	-	-

Knowledge Chunks Table

Column	Type	Sortable	Searchable	Toggleable	Special Features
Content	Text	-	✓	-	Limited to 100 chars
Status	Badge	✓	-	-	Color-coded with icon
Source	Text	-	✓	✓	Document reference
Created At	DateTime	✓	-	✓	-

Available Filters

Status Filter: Select by syncing, synced, or failed

Record Actions (Per Row)

Resource	Actions
Documents	Process, Chat, Download, View, Edit
Users	View (slide-over), Edit (slide-over), Delete
Roles	Edit, Delete
Chunks	Edit (slide-over), Delete

Header Actions (Table-Level)

Resource	Action	Description
Documents	Process All	Process all pending/failed documents
Chunks	Search & Replace	Bulk find/replace across all chunks

Bulk Actions

Resource	Available Bulk Actions
Documents	Delete (via Bulk Action Group)
Users	Delete (via Bulk Action Group)
Roles	Delete (via Bulk Action Group)
Chunks	Delete (via Bulk Action Group)

Security & Identity

Authentication Features

Two-Factor Authentication (2FA): (Currently Disabled) Optional TOTP-based secondary verification for enhanced security.
Session Management: View and logout of active sessions on other devices.
Email Verification: Required email verification workflow for new users.
Password Reset: Secure password recovery via email tokens.

Authorization System

Token-Based Auth: Secure API access via Laravel Sanctum with role-based abilities.
Role-Based Access Control (RBAC): Comprehensive permission system via Filament Shield.
Granular Permissions: Fine-grained control over view, create, update, delete, restore, and force-delete actions.
Custom Roles: Create custom roles with specific permission sets for different user types.
Guard System: Separate guards for web (Filament) and sanctum (API) authentication.

Security Best Practices

UUID Identifiers: Non-sequential identifiers for documents and sensitive resources.
Cascade Deletion: Automatic cleanup of related data to prevent orphaned records.
Rate Limiting: Configurable API throttling to prevent abuse.
Environment Isolation: Separate configurations for development, testing, and production.
Password Hashing: BCrypt hashing for all user passwords.
CSRF Protection: Built-in CSRF token validation for state-changing requests.

Support & Policy

📋 Item Support Policy

I am committed to providing high-quality support for the AI RAG Wrapper. Please review the following guidelines to understand what assistance is available.

✅ What's Included in Support

Answering technical questions about the item's features
Assistance with reported bugs and issues
Help with the included Installation Wizard
Configuration guidance for AI providers
API integration assistance
Troubleshooting document processing issues

❌ What's NOT Included

Customization services or adding new features
Installation on your server (available as paid service)
Issues caused by 3rd party plugins
Hardware or server limitations
Hosting-specific configuration (cPanel, VPS setup)
PHP/MySQL server configuration

🕐 Support Hours & Response Time

GMT+6 (Khulna, Bangladesh) 24 Hours Response Sunday - Thursday

I am based in Khulna, Bangladesh (GMT+6). I usually respond within 24 hours during business days (Sunday - Thursday). Messages received on Friday-Saturday (weekend in Bangladesh) will be addressed on Sunday.

📞 Contact Support

📧 Email: support@virtualmind.top

⏱️ Response Time: 24 hours during business days

⚠️ Before Opening a Support Ticket

Please ensure you have completed the following steps to help us assist you faster:

Read the documentation - All features and setup steps are documented
Check the Troubleshooting section - Common issues have documented solutions
Verify system requirements - PHP 8.3+, MySQL 8.0+, required extensions
Ensure queue worker is running - php artisan queue:work must be active
Check error logs - Review storage/logs/laravel.log for errors

When Contacting Support, Include:

Your Envato Purchase Code
PHP version (run: php -v)
Error logs from storage/logs/laravel.log
Steps to reproduce the issue
Screenshots if applicable
Your server environment (Shared hosting / VPS / Local)

💡 Tip: Tickets with complete information receive faster responses. Incomplete tickets may delay support.

💼 Paid Services Available

Need help beyond standard support? I offer the following paid services:

Server Installation - Full setup on your hosting/VPS
Custom Modifications - Feature additions or UI changes
Third-party Integrations - Connect with your existing systems
Priority Support - Expedited response (4 hours)

Contact me for a custom quote based on your requirements.

🙏 Third-Party Credits & Licenses

This project makes use of various third-party libraries, frameworks, and assets. We are grateful to the amazing developers behind these tools. Below is a complete list with attribution links and license information.

🌐 Core Framework & Platform

Project	Version	License	Link
Laravel	12.x	MIT License	laravel.com
PHP	8.3+	PHP License 3.01	php.net

🎨 Admin Panel & UI Components

Project	Version	License	Link
Filament	5.x	MIT License	filamentphp.com
Livewire	4.x	MIT License	livewire.laravel.com
Flux UI	2.9+	Commercial License	fluxui.dev
Tailwind CSS	4.x	MIT License	tailwindcss.com

🔐 Authentication & Security

Project	Version	License	Link
Laravel Sanctum	4.x	MIT License	GitHub
Laravel Fortify	1.x	MIT License	GitHub
Filament Shield	4.1+	MIT License	GitHub

🤖 AI & RAG Engine

Project	Version	License	Link
Laravel Markdown RAG	0.2+	MIT License	GitHub
PDF Parser	2.12+	LGPL-3.0	pdfparser.org
Laravel AI	Latest	MIT License	GitHub

📊 Monitoring & Utilities

Project	Version	License	Link
Filament Jobs Monitor	4.1+	MIT License	GitHub
Laravel Pail	1.2+	MIT License	GitHub
Laravel Boost	2.0+	MIT License	GitHub

🧪 Testing & Development

Project	Version	License	Link
Pest PHP	4.x	MIT License	pestphp.com
Pest Plugin Laravel	4.1+	MIT License	GitHub
Laravel Pint	1.24+	MIT License	GitHub
Faker	1.23+	MIT License	fakerphp.github.io
Mockery	1.6+	BSD-3-Clause	docs.mockery.io
Newman	6.1+	MIT License	GitHub

🔧 Build Tools

Project	Version	License	Link
Vite	7.x	MIT License	vite.dev
Laravel Vite Plugin	Latest	MIT License	GitHub
Autoprefixer	10.4+	MIT License	GitHub
Axios	1.7+	MIT License	axios-http.com
Concurrently	9.0+	MIT License	GitHub

✨ Fonts & Icons

Project	Version	License	Link
Inter Font	Latest	SIL Open Font License 1.1	rsms.me/inter
Outfit Font	Latest	SIL Open Font License 1.1	outfit.style
Heroicons	Latest	MIT License	heroicons.com

📦 Additional Dependencies

The following packages are also used under their respective open-source licenses:

Laravel Sail - MIT License
Laravel Tinker - MIT License
Livewire Blaze - MIT License
Tailwind CSS Oxide - MIT License
LightningCSS - MPL 2.0
Rollup - MIT License

📜 License Summary

Most dependencies in this project use permissive open-source licenses:

MIT License

Permissive, allows commercial use, modification, distribution, and private use.

Apache 2.0

Permissive with patent rights and attribution requirements.

SIL OFL 1.1

Free font license, allows commercial use and modification.

LGPL-3.0

Allows linking to proprietary software with dynamic linking.

📝 Note: All trademarks, logos, and brand names are the property of their respective owners. Their use does not imply endorsement or affiliation.

Changelog

Version 1.0.0 🚀 - Initial Release

Release Date: 2026-04-02

Core Features

Complete RAG pipeline with smart chunking and vector embeddings
Multi-provider AI support (12+ LLM providers)
Filament v5 admin dashboard with monitoring widgets
RESTful API v1 with SSE streaming
Document management with background processing
Role-based access control via Filament Shield
Personal Access Token Management Resource

Pipeline Optimizations

Semantic caching for reduced API costs
Optional reranking for improved precision
Query rewriting for better retrieval
Exponential backoff retry logic

Developer Tools

Postman collection with automated tests
Interactive API documentation
Comprehensive test suite with Pest 4
Laravel Pint code formatting

Introduction

What is RAG?

🚀 Quick Start

5-Minute Quick Start

Key Features

🤖 AI & RAG Engine

Smart Chunking

Semantic Search

Reranking Pipeline

Query Rewriting

Semantic Caching

Streaming Responses

Exponential Backoff Retry

Chat Rate Limiting

📄 Document Management

Multi-format Support

Automatic Text Extraction

Background Processing

Real-time Progress

Custom Instructions

UUID Identification

Cascade Deletion

⚙️ Technology Stack

Laravel 12

Filament v5

Livewire 4

Flux UI

Pest 4

Tailwind CSS 4

System Requirements

Installation Guide

Standard Installation

What Happens During Installation

License Verification Behavior

Key Design Principles

Detailed cPanel Installation Guide

Creating Database using Cpanel

Uploading Files using Cpanel

Renaming .env.example to .env

Uploading Files using FTP

Background Processing & Scheduled Tasks (Crucial)

Recommended: Laravel Scheduler (Single Cron Job)

Alternative: Direct Queue Worker

VPS (Supervisor for Continuous Processing)

Shared Hosting (Cron Job)

Installation Wizard Steps

Step 1: Preflight Checks

Step 2: License Verification

Step 3: Database Configuration

Step 4: Application Configuration

Step 5: Create Super Admin

Installation Complete!

🔧 Troubleshooting

Common Issues

Issue Ingestion stuck at 0%

Issue 404 errors on API routes

Issue PDF parsing fails

Issue "Class not found" errors

Issue Permission denied errors

Issue Email verification not sending

Issue AI responses are slow

Getting Help

Admin Dashboard (Filament)

Navigation Structure

Dashboard Widgets

Management Resources

👥 Users Resource

Form Inputs (Create/Edit)

Table Columns

Available Actions

API Endpoints

🔐 Roles Resource (Shield)

Form Inputs (Create/Edit)

Table Columns

Available Permissions

API Endpoints

🔑 API Tokens Resource

Resource Features

Table Columns

Available Actions