MemroOS Setup Guide

Repo: lac5q/memroos Stars: ⭐6 | Language: TypeScript

Overview

MemroOS is a memory operating system and governance layer for AI agents. It provides agent workflows, dispatch, proof, and context continuity — ensuring your agents don't lose track of what they're doing across sessions. Local-first, MCP-ready.

Key capabilities: - Agent memory with context continuity across sessions - Workflow dispatch and proof tracking - Multi-agent observability - Local-first architecture (no cloud dependency) - MCP server for any agent harness

Installation

git clone https://github.com/lac5q/memroos.git
cd memroos
npm install
npm run build

Configuration

{
  "server": { "port": 3001 },
  "storage": { "path": "~/.memroos" },
  "agents": ["hermes", "openclaw", "codex"]
}

Hermes Integration

Add to MCP config:

{
  "mcpServers": {
    "memroos": {
      "command": "node",
      "args": ["path/to/memroos/dist/server.js"],
      "env": { "MEMROOS_HOME": "~/.memroos" }
    }
  }
}

Pitfalls

• Early-stage project (6 stars) — APIs may change
• TypeScript runtime requires Node.js 20+