Installation

Engram is distributed as a self-contained local operating system. You can spin up the entire stack—Frontend, Backend, and Vector Database—with a single command.

1. Prerequisites

System Requirements

Node.js 18+ (for the dashboard interface)
Docker Desktop (to run the Python 'Brain' and Vector DB)
8GB RAM minimum (16GB recommended for Llama 3)

2. Run the Installer

Open your terminal and run the ephemeral installer. This pulls the latest Docker images and configures your local environment.

>npx engram-os@latest

Core

Downloads the Python Intelligence API.

Models

Fetches Llama 3 & Whisper weights (4GB).

CLI

Configures global aliases like g-commit.