# Introduction

> **The AI Inception Engine for the Mnemosyne Neural OS ecosystem.**

[![version](https://img.shields.io/badge/version-1.2.1-8B5CF6)](https://www.npmjs.com/package/@mnemosyne_os/forge) [![License](https://img.shields.io/badge/License-MIT-green)](https://github.com/yaka0007/Mnemosyne-Neural-OS/blob/main/cli/LICENSE/README.md) [![TypeScript](https://img.shields.io/badge/TypeScript-strict-blue)](https://www.typescriptlang.org/) [![Node.js](https://img.shields.io/badge/Node.js-22%2B-brightgreen)](https://nodejs.org/) [![ecosystem](https://img.shields.io/badge/ecosystem-Mnemosyne%20Neural%20OS-8B5CF6)](https://github.com/yaka0007/Mnemosyne-Neural-OS)

## What is MnemoForge?

**MnemoForge** is the official CLI for the [Mnemosyne Neural OS](https://github.com/yaka0007/Mnemosyne-Neural-OS) ecosystem.

It gives AI agents something they fundamentally lack: **persistent, versionable, IDE-agnostic memory**.

> *"Don't just scaffold code. Scaffold intelligence."*

## The Problem

Every time an AI agent starts a new session, it starts from zero. It forgets:

* What the package scope is (`@mnemosyne_os`, not just anything)
* That chronicles are written by the agent, never by the human
* The architectural decisions made in the last 3 sessions
* Why a certain rule exists

**MnemoForge solves this** with two systems:

| System         | Purpose                                                      |
| -------------- | ------------------------------------------------------------ |
| **Chronicles** | Structured memory files capturing key decisions and sessions |
| **Workspace**  | Project-level rules that any agent reads at session start    |

## Install

```bash
npm install -g @mnemosyne_os/forge
```

## Quick Start

```bash
# Initialize the vault
mnemoforge chronicle init

# Check workspace rules before starting work
mnemoforge workspace show

# Archive a chronicle written by your agent
mnemoforge chronicle archive --file "path/to/CHRONICLE-2026-04-05-my-decision.md"
```

## Core Concept

MnemoForge implements **Neural Coding** — a development methodology where:

* The **human** holds the intention and understands the system
* The **agent** understands the context and executes without mechanical instructions
* The **memory** persists outside of any IDE, any session, any conversation

The code follows the thought. Not the other way around.

***

*Part of the* [*Mnemosyne Neural OS*](https://github.com/yaka0007/Mnemosyne-Neural-OS) *ecosystem.*