Metadata-Version: 2.4
Name: nebulix
Version: 1.0.4
Summary: Nebulix AI Framework with split Local (Offline) and Cloud (Gist) engines.
Home-page: https://github.com/NebulixLabs
Author: Kshitij Rajput
Classifier: Programming Language :: Python :: 3
Classifier: License :: Other/Proprietary License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.7
Description-Content-Type: text/markdown
Requires-Dist: requests>=2.25.0
Requires-Dist: fastapi>=0.65.0
Requires-Dist: wikipedia-api>=0.5.4
Requires-Dist: sympy>=1.8
Requires-Dist: rapidfuzz>=2.0.0
Requires-Dist: opencv-python>=4.5.0
Requires-Dist: numpy>=1.19.0
Requires-Dist: pytesseract>=0.3.8
Requires-Dist: Pillow>=8.2.0
Requires-Dist: googletrans==4.0.0-rc1
Requires-Dist: beautifulsoup4>=4.9.0
Dynamic: author
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# 🧬 Nebulix AI Framework (v1.0.3)

**Nebulix AI** is a professional, high-performance, modular AI framework developed by **Nebulix Labs**. It features a split architecture offering both a 100% **Local (Offline)** heavy knowledge engine and a **Cloud (Online)** fallback framework.

* **License Holder**: Kshitij Rajput (Founder of Nebulix Labs)
* **License Type**: Proprietary & Close Source

---

## ⚠️ CRITICAL DEVELOPER GUIDELINES (Read Before Integration)

### 1️⃣ Run Integration Tests First
Before deploying or connecting this framework to your production frontend, **ensure you test all modules (`local` and `cloud`) locally** using dummy sessions to verify optimal connectivity.

### 2️⃣ Understand the Unique Token Requirement (Strict Rule)
Do NOT attempt to use this framework without a valid authorization token, even for normal text processing or local offline actions.
* **Authentication is mandatory** for all modes.
* If you bypass or skip the `auth_token` parameter during initialization, critical modules (like Robot Mode, Vision Processing, and custom responses) will be completely locked or missed. 
* Always enforce token passing in your production pipeline.

### 3️⃣ Frontend Parsing & Unique Output Patterns
The framework returns responses embedded with specialized structural markers to safely bypass large LLM dependencies. Your frontend team **must analyze these unique text patterns and custom symbols** to accurately render headers and code blocks:

* **Headers Indicator (`****`)**: When a response block begins or ends with `****` (e.g., `****Header Text****`), the frontend must treat this text as a structural **Header** component.
* **Code Block Indicator (`*`)**: Raw code segments generated by the framework (Python, HTML, React) are explicitly wrapped inside four asterisks (`*`) at the beginning and the end. 
  * *Example format returned by engine:* `* \n [Your Generated Code Here] \n *`
  * The frontend must capture text inside these unique markers to dynamically inject them into an Editor or a Highlight Component.

---

## 🛠️ Quick Architecture Implementation

### 📂 Mode A: 100% Local Heavy Engine (Offline data)
Loads `perfect_1.json` through `perfect_46.json` dynamically from local storage while verifying the core developer signature online.

```python
import asyncio
from nebulix import local  # Imports the offline engine package route

async def run_offline():
    # Auth token is strictly required for normal use too
    ai = local(auth_token="PROPRIETARY_AUTH_TOKEN", power_level=4, ai_name="My AI")
    
    response = await ai.chat(token="user_session_id", query="hi")
    print(response["answer"])

asyncio.run(run_offline())
