Skip to content

Quantum Sentinel – A modular WordPress AI assistant that defaults to free Groq and can instantly switch to xAI, OpenAI, or Google (Gemini) APIs while tapping your on-site knowledge base.

License

Notifications You must be signed in to change notification settings

ResearchForumOnline/AgentZeroWordPress

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🚀 Quantum Sentinel

Quantum Sentinel Avatar

If you would like an AI chatbot that you can edit and re-brand without coding try our other plugin, though it is not using the advanced frameworks this Sentinel plugin is using. Our other WordPress plugin fully customizable plugin: https://github.com/ResearchForumOnline/WordpressAIchatbot

Modular AI assistant for WordPress — free ⚡ Groq by default, with effortless switches to xAI Grok, OpenAI GPT-4o/3.5 or Google Gemini, plus instant knowledge-base search.

WordPress 6.x PHP 7.4+ Groq API License: MIT


📚 Table of Contents


✨ Features

🔸 Free Groq backend Llama-3-70B at zero cost with sub-20 ms latency
🔌 Pluggable engines Swap to xAI Grok, OpenAI GPT-4o / 3.5, or Google Gemini in one click
📖 Knowledge-base aware Inject posts, pages, or custom JSON snippets so answers always match your content
🛡 Privacy controls Optional message logging, IP anonymisation, configurable data-retention
🛠 Developer-friendly 30 + hooks, PSR-12 PHP, concise vanilla-JS front-end, TypeScript typings
Lazy asset loading Styles and scripts enqueue only on pages containing the shortcode or block
🌍 i18n ready All strings wrapped in __() / _e() for translation

🏗 Architecture

quantum-sentinel/

├── quantum-sentinel.php # Main loader & constants

├── admin/

│ └── quantum-sentinel-admin.php

├── includes/

│ ├── quantum-sentinel-core.php

│ └── class-qs-client.php # Engine abstractions (Groq, xAI, OpenAI, Gemini)

├── assets/

│ ├── quantum-sentinel-avatar.png

│ ├── qs-chat.css

│ └── qs-chat.js

└── languages/

└── quantum-sentinel.pot

All engine endpoints extend QS_Client_Base, so adding a new provider is a 50-line subclass.


📦 Requirements

Item Minimum
WordPress 5.8 (tested up to 6.5.0)
PHP 7.4 (fully tested on 8.3)
Browser Fetch & EventSource support (all modern browsers)
API Key Free Groq key or xAI / OpenAI / Gemini key(s)

🛠 Installation

# 1. Download latest release
wget https://github.com/ResearchForumOnline/AgentZeroWordPress/releases/latest/download/quantum-sentinel.zip

# 2. Upload via WP-Admin or unzip into wp-content/plugins/
unzip quantum-sentinel.zip -d wp-content/plugins/

# 3. Activate
wp plugin activate quantum-sentinel   # or use WP-Admin → Plugins

# 4. Configure
#   WP-Admin → Settings → Quantum Sentinel → enter API key(s)
⚙️ Quick Configuration
Setting	Default	Description
AI Engine	groq	Choices: groq, xai, openai, gemini
Model	llama3-70b-instruct	Engine-specific slug
Temperature	0.7	0 = strict, 1 = creative
Max Tokens	1024	Hard cap per reply
KB Source	internal	internal, posts, both
Log Retention	0	0 disables logging

🚀 Usage
Gutenberg Block
Add the “Quantum Sentinel” block.

Configure avatar side, theme (light, dark, auto) and KB subset.


[quantum_sentinel avatar="right" theme="dark" kb="product_faqs"]
PHP Helper
php
Copy
Edit
if ( function_exists( 'quantum_sentinel_render' ) ) {
    echo quantum_sentinel_render( array(
        'theme'   => 'auto',
        'avatar'  => 'left',
        'engine'  => 'openai',
        'model'   => 'gpt-4o-mini',
    ) );
}
🔧 Hooks & Filters
Hook	Context
qs_ai_request_args	Filter final request payload before HTTP call
qs_system_prompt	Replace or augment the default system prompt
qs_kb_items	Inject custom KB array programmatically
qs_after_reply	Fires after each successful response ($chat_id, $reply)
qs_user_avatar_url	Override the default avatar PNG

Full list in docs/hooks.md.

❓ FAQ
<details> <summary><strong>Is Groq really free?</strong></summary>
Yes. As of July 2025 Groq offers free access to Llama-3-70B for hobby and production use.
Check their dashboard for current rate-limits before launch day.

</details> <details> <summary><strong>Can I self-host the LLM?</strong></summary>
Absolutely. Implement a QS_Client_MyLocalLLM subclass that returns OpenAI-compatible JSON and hook it via qs_client_factory.

</details> <details> <summary><strong>Does Quantum Sentinel store user data?</strong></summary>
Message logs are off by default. When enabled, you choose retention days (0 = none) and whether IPs are anonymised. All data lives in your own DB—no SaaS middle-man.

</details>
🗺 Roadmap
 Server-Sent Events streaming

 Vision endpoints (Gemini Pro Vision, GPT-4o-Vision)

 Optional SQLite + pgvector vector store

 Full React admin panel

 Multisite (network-wide) settings

🤝 Contributing
Fork → git switch -c feature/your-feature

composer install && npm install

Run linters & tests: composer lint, npm run lint

Pull Request with a clear description — GH Actions runs PHPCS, PHPUnit & Vitest

After merge, your name appears in AUTHORS.md 🎉

📜 License
Quantum Sentinel is released under the MIT License, which is GPL-compatible.
See LICENSE for the full text.

About

Quantum Sentinel – A modular WordPress AI assistant that defaults to free Groq and can instantly switch to xAI, OpenAI, or Google (Gemini) APIs while tapping your on-site knowledge base.

Resources

License

Stars

Watchers

Forks

Packages

No packages published