Skip to content
Conduit

mcp server

NEURA.AI

NEURA - A ARM based AI OS -Welcome to the new era of personal assistance

ved3010/NEURA.AIcommunitylistedunreviewed
Use tool

Set up

Runs on your computer. Pick your agent and operating system.

Before you start

  • uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`)
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "neura-ai": {
      "command": "uvx",
      "args": [
        "neura-ai"
      ]
    }
  }
}

same format Cursor, GitHub Copilot CLI, Junie, JetBrains AI Assistant, Augment (Auggie CLI), Factory Droid, Kiro, Devin CLI +12

Values in angle brackets are placeholders. Restart Claude Desktop after editing its config.Claude Desktop docs

README

synced Oct 22, 2026

Neura AI — Voice & Holographic AI Assistant

Project Overview

Neura AI is a next-generation AI assistant that combines a real-time voice pipeline with an interactive 3D holographic orb interface. It is interesting because it demonstrates how advanced AI models and spatial interfaces can run efficiently on Arm64 environments (Apple Silicon macOS), combining high-performance computing with an intuitive user experience.

Functionality / Output

The project provides a fully functional desktop assistant featuring:

  • 3D Holographic UI: A WebGL-based interface built with Next.js and Three.js, incorporating MediaPipe hand tracking for gesture control.
  • Voice Agent: A LiveKit voice pipeline that utilizes a powerful LLM for reasoning and OpenAI TTS for speaking back in real time.
  • FastMCP Server: Exposes system tools (news, web search, system info) to the voice agent.

Setup Instructions

The application is designed to run seamlessly on Arm-powered devices like Apple Silicon Macs (Arm64 environment). Please follow the frontend and backend setup sections below to build, run, and validate the project locally on your machine.

Repository Components

  1. Interactive Frontend (Neura AI Hologram UI): Next.js + Three.js holographic orb desktop widget.
  2. Voice Backend (Neura AI Core): LiveKit voice pipeline agent and FastMCP server.

1. Interactive Frontend (Neura AI Hologram UI)

An Iron Man–inspired holographic orb built with Next.js, Three.js, and MediaPipe hand tracking — control it with your bare hands through your webcam.

🔮 This is the open-source interface of ULTRON — my AI that talks in real time and controls Android devices by itself. Read the write-up or the X post 📱 Watch the demo on Instagram

Getting started (Frontend)

npm install
npm run dev

Open http://localhost:3000.

Controls

Mouse / touch

InputAction
DragSpin the orb
Scroll / pinchZoom in & out

Hand gestures (webcam)

Click GESTURES OFF (or press G) and allow camera access, then:

GestureAction
Pinch (thumb + index) one hand and move itSpin the orb
Pinch with both hands, spread apart / bring togetherZoom in / out

Keyboard

KeyAction
GToggle hand gestures
RReset the view
+ / −Zoom in / out

2. Voice Backend (NEURA.AI — Tony Stark Demo)

🎉 Official Public Release: E.V. is now officially released to the public as a standalone application! You can easily install it without needing to set up the development environment.

"Fully Responsive Intelligent Digital Assistant for You"

A Tony Stark-inspired AI assistant split into two cooperating pieces:

ComponentWhat it is
MCP Server (uv run friday)A FastMCP server that exposes tools (news, web search, system info, …) over SSE. Think of it as the Stark Industries backend — it does the actual work.
Voice Agent (uv run friday_voice)A LiveKit Agents voice pipeline that listens to your microphone, reasons with an LLM (Gemini 2.5 Flash by default), and speaks back with OpenAI TTS — all while pulling tools from the MCP server in real time.

Demo: Instagram reel

How it works (Backend)

Microphone ──► STT (Sarvam Saaras v3)
                    │
                    ▼
             LLM (Gemini 2.5 Flash)  ◄──────► MCP Server (FastMCP / SSE)
                    │                              ├─ get_world_news
                    ▼                              ├─ open_world_monitor
             TTS (OpenAI nova)                     ├─ search_web
                    │                              └─ …more tools
                    ▼
             Speaker / LiveKit room

The voice agent connects to the MCP server via SSE at http://127.0.0.1:8000/sse (auto-resolved to the Windows host IP when running inside WSL).

Quick start (Voice Backend - For Developers)

1. Prerequisites

  • Python ≥ 3.11
  • uv — run pip install uv or curl -Lsf https://astral.sh/uv/install.sh | sh
  • A LiveKit Cloud project (the free tier works)

2. Install dependencies

uv sync

(This creates the .venv and installs all dependencies)

3. Set up environment

cp .env.example .env

(Open the newly created .env file and fill in your API keys using the reference below)

4. Run — two terminals

Terminal 1 — MCP server (must start first)

uv run friday

Starts the FastMCP server on http://127.0.0.1:8000/sse. The voice agent connects here to fetch its tools.

Terminal 2 — Voice agent

uv run friday_voice

Starts the LiveKit voice agent in dev mode — it joins a LiveKit room and begins listening. Open the LiveKit Agents Playground and connect to your room to talk to FRIDAY.

License

MIT

↑↓ navigate openesc close