RiMM.

A small agent on each Windows PC and Mac reports inventory and health to one dashboard, organised by business, site and group — with scripts, a remote terminal, schedules and alerts to follow.

What it will be

RiMM is a remote monitoring and management (RMM) tool for looking after other people’s computers. A small agent runs on each Windows PC or Mac and reports to one dashboard, where devices are organised by business, site and group. The first stage is built and tested on a real Windows 11 PC; the rest follow stage by stage.

The plan

1. Enrollment and inventory — built: a Windows PC joins its group with a short-lived, limited-use install token and reports its hardware and operating system.
2. macOS and full inventory — network adapters, installed apps, signed-in users, and live and historical metrics.
3. Scripts and fleet jobs — versioned scripts with typed parameters, run across many devices with safe concurrency limits.
4. Remote terminal — a shell on any device from the browser, audited and never recorded.
5. Schedules — recurring jobs with time zones and maintenance windows.
6. Alerts and webhooks — threshold alerts with acknowledge and resolve, and signed webhooks.
7. Automatic updates — agent updates from a signed manifest, with a health check and rollback.
8. Deployment review — tested backups and restores, packaging, and a 100-agent load test.

Design principles

• Agents only call out — each agent opens an outbound, encrypted connection; nothing on the device listens for the server.
• The server decides — a device’s business, site and group come from the server, never from what the agent claims, and everything an agent reports is treated as untrusted text.
• Learning from real breaches — the design draws on open-source RMMs (MeshCentral, Tactical RMM) and on what went wrong in the Kaseya, ScreenConnect and SimpleHelp incidents.
• Google sign-in only — the admin console keeps no passwords of its own.