Monitor · Alert · Stay ahead

Less Downtime.
More Tomorrow.

Monitoring & on-call for what matters. Built for whoever operates their own servers and would rather spend the night sleeping.

Open source. No sign-up form here: HopMango is installed by talking to your assistant, or by hand from the docs.

Observe · Respond · Improve

Reliable systems, happier people.

Three decisions we made on day one, and wrote down so they stay made.

Evaluation at the edge

The agent decides on the host whether something is wrong. Only state changes and aggregated numbers travel to the center: no raw metrics, no ingestion bill, no surprises.

ADR-0005

An agent you can trust

No listening ports on any code path. Runs without root. Accepts only configuration signed for that host, and a stale bundle is refused. Each of these is a CI test, not a promise.

Agent security

The portal is optional

Everything you can click exists in the API and in MCP. Tell your assistant to instrument a project and it deploys agents, adopts checks and routes alerts, without a single screen.

ADR-0011

One mango, every shift

Hop into uptime!

Hosts, checks, synthetics, browser sessions and the people on call, in one place.

Monitor

Hosts, checks, synthetics and RUM in one panel, with a time range you can share as a link.

Alert

Every rule carries a duration. Nothing pages you for a blip, and a check that flaps gets quiet before it gets loud.

On-call

A native iOS app that wakes you when it matters, and lets you acknowledge from the lock screen.

All good

Status at a glance: what is up, what is degraded, and what nobody has picked up yet.

Fixed!

Incidents with context. Mango AI proposes a likely cause; it never touches your systems.

Coffee time

Configuration lives in a file or in the UI, your choice, and the two never clobber each other. Quiet nights follow.

Good systems, brighter days

A dashboard that gets out of the way

Light or dark, one shared time range, charts that keep up with real infrastructure. And when there is nothing wrong, it says so in three words.

  • Incidents first: that is the unit a person works with.
  • Every screen shows the API and MCP call behind it.
  • Empty states teach you how to enrol the first host.

Mango On-Call

For the four in the morning

A notification that stays silent in a focus mode is an incident log, not on-call. HopMango ships a native iOS app so the alert that matters gets through, and the acknowledge is one tap away.

How the app works