VORA Core CLI

vora-ai is the local CLI and gateway core for VORA, Zah's voice-first AI agent project.

This package provides the parts that run on the user's machine today:

• the vora CLI
• the local gateway/runtime
• onboarding and configuration flows
• model/provider setup
• local workspace + agent state

Phase 0 focus is simple local setup, stable defaults, and a clean base for the VORA desktop/app layers that will sit on top of this core.

Install

Requirements:

• Node.js 22.14+

Install from npm:

npm install -g vora-ai@latest

One-liner (Mac/Linux):

curl -fsSL https://heyvora.fun/install.sh | bash

One-liner (Windows - PowerShell):

iwr -useb https://heyvora.fun/install.ps1 | iex

Install from a local tarball:

npm install -g ./vora-ai-<version>.tgz

Recommended first run:

vora onboard

Gateway Default

VORA now uses this local gateway port by default:

ws://127.0.0.1:27106

Start it manually if needed:

vora gateway --port 27106

For a background service, use:

vora onboard --install-daemon

Model Setup

Current setup flows in this package include:

• OpenAI / OAuth-backed flows when configured
• local/free Ollama setup
• custom OpenAI-compatible or Anthropic-compatible providers

Ollama (local/free)

If you want a free local model:

ollama serve
ollama pull qwen3:4b
vora onboard

Then choose Ollama (Local/Free) in the model/auth setup step.

VORA will configure a local ollama provider in your model config and use the model already pulled on your machine.

Useful Commands

vora onboard
vora gateway --port 27106
vora configure
vora doctor
vora models

Scope

This npm package is the core runtime and CLI, not the final end-user desktop shell.

Voice UX, wake-word UX, and the planned Agora/Tauri product layers are separate layers built on top of this core.

Repository

• GitHub: https://github.com/vora-ai/vora-core
• License: MIT