DeadWatch · How it works

From a failed check to your phone, before anyone notices.

Add your sites. DeadWatch runs every check on your schedule, confirms what's actually broken, and pings you on the channel you chose. Here's the whole path — at a glance.

01
Check
Distributed workers, multiple regions
02
Ingest
Results collected & stored
03
Score
Graded, classified by reason
04
Correlate
Incident opened & de-duplicated
05
Notify
Email · Slack · Discord · Webhook
06
Resolve
Open → ack → resolve
07
Publish
Status page reflects it
01 · The watch

Every check runs from more than one place.

DeadWatch runs each monitor on your schedule — uptime, speed, SSL, DNS, visual changes and the rest — from more than one region. So a single hiccup on one server never gets mistaken for a real outage.

/watch — fleet
DeadWatch fleet dashboard showing many monitored sites and their live health at a glance
One live health picture across every site you watch.
02 · The filter

Only real problems reach you.

A single blip is ignored. DeadWatch confirms a failure before it pages anyone, folds related failures into one incident — so a big outage is a single alert, not fifty — and quiets a flapping site down to one thing to look at instead of a scroll of noise.

/watch/site — incident
Site detail view showing a confirmed incident with the uptime timeline
One confirmed incident, timelined — not a wall of duplicate alerts.
03 · The alert

It reaches you on the channel you chose.

When something real happens, the alert goes where you want it — Email, Slack, Discord, or a webhook into your own tools. Paste a Slack or Discord link and it just works. Route different clients to different places, and trust it to arrive — delivery is durable, so an alert never quietly vanishes.

Slack Discord Generic webhook
/watch/notifications
Notification settings showing Email, Slack, Discord and generic webhook channels
Four channel types, wired how you like — per site or per client.
04 · After the alert

Open, acknowledged, resolved.

Acknowledge an incident and the repeats quiet down while you work it. When the site recovers, DeadWatch marks it resolved. Planned deploy? A maintenance window holds the alerts — the checks keep running, they just don't page.

Stage 1
Open
A confirmed problem opens an incident and the first alert goes out.
Stage 2
Acknowledged
Someone owns it — repeats quiet down while it's worked.
Stage 3
Resolved
Recovery is detected and the incident closes.
See it end to end

Point it at your sites. Watch the pipeline work.

Add a roster and DeadWatch starts running the full path — check, score, correlate, notify — from the first interval. Explore what each stage watches, or start now.