Hugging Face's logo

Spaces:
neural-arun
/
ArunCore
Runtime error

App Files Community
ArunCore / data /github /aruncore /README.md
GitHub Actions Bot
deploy: automated sync from github main
985f3ee
preview code
|
Raw
History Blame Contribute Delete
4.87 kB
metadata 
project_name: aruncore
github_url: https://github.com/neural-arun/aruncore
language: TypeScript
stars: 1
topics:
  - None
updated_at: 2026-08-08T22:16:43.000Z

aruncore

GitHub Repository: https://github.com/neural-arun/aruncore
Primary Language: TypeScript | Stars: 1 | Forks: 0
Description: No description provided.

title: ArunCore AI Assistant & Multi-Tenant Enterprise Engine emoji: 🧠 colorFrom: green colorTo: blue sdk: docker app_port: 7860 pinned: false

🧠 ArunCore β€” AI Systems Architect & Multi-Tenant Enterprise Engine

ArunCore is a production-grade, stateful, agentic portfolio and multi-tenant AI Assistant platform built for Arun Yadav (AI Systems Architect specializing in Healthcare & Education).

It features a 100% Data-Driven Multi-Tenant Enterprise JSON Engine that can instantly spin up custom 24/7 AI Course Advisors and Sales Representatives for any tutor, instructor, or consultant (e.g. Ed Donner) simply by passing a query parameter (?tutor=ed_donner) backed by a single 238-key JSON schema.

🌟 Key System Features

  1. πŸ€– Multi-Tenant Enterprise JSON Engine (?tutor=<tutor_id>):

    • Dynamically loads tutor profiles, real course catalogs, bios, social links, FAQs, and custom system prompts from data/<tutor_id>_enterprise_dictionary.json.
    • Zero Code Changes Needed: To onboard a new tutor or client, duplicate demos/general.json, fill their bio/courses/colors, and open ?tutor=<new_client_id>.

  2. 🎨 Dynamic CSS Brand Theme System:

    • Reads theme_design_system from the active JSON file and injects custom CSS variables (--accent-green, --accent-green-hover, --border-accent, --bg-accent-soft).
    • Automatically adapts the entire UI (buttons, active tabs, badges, quick questions, sidebar status dots) to the tutor's brand personality.
    • Defaults seamlessly to Arun Yadav's original sacred Emerald Green (#10b981) portfolio when no parameter is provided.

  3. πŸŽ“ Flagship Ed Donner Integration Demo (?tutor=ed_donner):

    • Complete real-world course catalog covering Ed's flagship tracks on Udemy (AI Engineer Core Track, Agentic Track with MCP, Production MLOps Track, AI Builder, AI Coder, AI Leadership).
    • Custom Deep Teal (#0d9488) brand theme, local high-res avatar, and clean SVG social symbols (πŸŽ“, πŸ’Ό, 𝕏, 🌐).

  4. ⚑ 100% Automated Telegram Alerts (Zero LLM Dependency):

    • Automatically triggers instant notifications to the instructor's Telegram Alert Bot for visitor messages.
    • Includes: User Question, AI Response, Session ID, and a 1-Click Magic Join Link.

  5. πŸ‘¨β€πŸ’» 1-Click Magic Link 3-Way Real Human Takeover:

    • Tapping the 1-click link in Telegram opens the website in Admin Mode on any device.
    • Unlocks a dedicated input box for the real human instructor to chat live alongside the AI in a 3-Way Conversation Room.

  6. 🧠 AI Agent Onboarding Guide & SOP (demos/TUTOR_ONBOARDING_GUIDE.md):

    • A phase-by-phase protocol for AI Coding Agents and developers to research and populate new tutor JSON files with strict Zero-Hallucination Guardrails.

πŸ“ Directory Structure & Key Files 

aruncore/
β”œβ”€β”€ core/
β”‚   β”œβ”€β”€ agent.py               # Dynamic Tutor LLM agent loader & prompt builder
β”‚   └── api.py                 # FastAPI backend server with /chat & /api/config endpoints
β”œβ”€β”€ data/
β”‚   β”œβ”€β”€ arun_enterprise_dictionary.json     # Arun Yadav master JSON schema
β”‚   └── ed_donner_enterprise_dictionary.json # Ed Donner lead JSON schema
β”œβ”€β”€ demos/
β”‚   β”œβ”€β”€ general.json           # Blank master JSON schema template
β”‚   β”œβ”€β”€ TUTOR_ONBOARDING_GUIDE.md # AI Agent Onboarding Protocol SOP
β”‚   └── ed_donner_enterprise_dictionary.json # Gold standard reference example
β”œβ”€β”€ frontend/
β”‚   β”œβ”€β”€ app/page.tsx           # Main Next.js page with URL tutor search param fetch
β”‚   β”œβ”€β”€ components/            # Header, Sidebar, ChatPanel, ProjectsView, ManifestoView, HandoffModal
β”‚   └── public/                # Local avatars (ed_donner.png, profile_photo.png)
└── PRODUCTION_CLIENT_ONBOARDING_PLAYBOOK.md # Production deployment playbook for $300+ clients

πŸš€ Quick Start & Local Execution

1. Backend Server (FastAPI on Port 8000) 

source .venv/bin/activate
python3 -m uvicorn core.api:app --reload --port 8000

2. Frontend Dev Server (Next.js on Port 3000) 

cd frontend
npm run dev

3. Test URLs