Getting started

By the end of this page you will have an account, one HTTP monitor being probed from a WireWatch region, and an email address that receives an alert when that monitor goes down. It costs nothing — every step here fits inside the Free plan, whose limits are listed at the bottom.

Four steps, in order. Step 2 is the one that decides whether an alert ever reaches you, so do not skip it.

1. Create an account

Go to app.wirewatch.io/signup and give three things: an organisation name, your email address and a password.

Signing up creates the organisation, makes you its Owner, and signs you in. You land on the monitor list.

An organisation is the tenancy boundary: monitors, incidents, channels, status pages and agents all belong to exactly one organisation, and nothing crosses between them.

2. Verify your email address

WireWatch emails you a verification link when you sign up. Until you open it, a banner sits at the top of every dashboard page telling you the address is unverified. Its Resend button sends another link, and so does the same control in Settings.

The link is deliberately usable without a session — mail is routinely read in a different browser from the one you signed up in — so you can open it wherever it lands.

This step is not paperwork. Confirming your address is the moment WireWatch gives your organisation its alerting:

• a notification channel named Default email ( … ), of type email, pointed at the address you just confirmed;
• a notification policy named Default policy ( … ) with a single level, no delay, and no severity filter — it notifies on every incident;
• that policy is recorded as the organisation’s default.

Nothing else is required to receive alerts.

Does a new monitor alert anyone with no further setup?

Yes — a new monitor alerts you with no channel setup.

When an incident opens, WireWatch looks for the policy to use in this order:

monitor → service → organisation default

A monitor with no policy of its own, in a service with no policy of its own, falls through to the organisation default seeded above. So the monitor you create in step 3 is already wired to your inbox.

The one way to end up silent is to never confirm your address. The seed runs at confirmation precisely because an alert may only ever be sent to an address someone has proved they can read — an unconfirmed address is never mailed.

Incidents that predate the seed are not replayed

Confirming your address does not mail you the incident history the organisation already had. Those events are marked as already handled at the moment the channel is created, so the first alert you receive is for the first incident after it.

3. Create your first monitor

In the dashboard, open Monitors and press + New. On the Free plan the type is http; the form asks for:

• a name — anything; it is what the alert email will call it;
• a URL — an absolute http(s) URL. It must be publicly reachable: a private, loopback or link-local target is refused at creation time, because a WireWatch region cannot reach inside your network. (That is what agents are for — see Probe locations.)
• a method — GET unless you need otherwise;
• an interval — how often to probe. Free allows 300 seconds;
• a timeout — must be shorter than the interval, in milliseconds;
• a probe location — Free allows one. Pick the region closest to your users;
• optionally, assertions — “the response must be 200”, “the body must contain ok”. With none, the check passes whenever the request completes without a transport error. See Assertions.

Save. The monitor is active immediately and the assignment reaches the probe fleet within a minute.

4. Read the first result, then the first alert

Open the monitor. Within one interval you will see a result per probe location: an up/down verdict, a response time, and the raw status code.

Nothing alerts on a single failure — that would page you for every dropped packet. The path from a broken target to your inbox is:

1. the probe fails, and keeps failing: a location must fail 2 checks in a row before it counts as down there;
2. because you are probing from one location, that verdict cannot yet be corroborated. WireWatch sends one extra probe from a different location — a region you are not using, or one of your own agents — and holds;
3. if the second opinion agrees, the incident opens. If it disagrees, the location is treated as locally broken and no incident opens;
4. the incident is matched against your policies, resolves to the organisation default from step 2, and the email is queued and sent.

Recovery is symmetric: 2 successful checks in a row per location, and the incident resolves and sends an all-clear.

If there is nowhere at all to ask for a second opinion, WireWatch falls back to time instead of space: 3 failures in a row on the lone location opens the incident by itself.

Every alert email carries a one-click link that turns that one channel off, so a runaway alert is always stoppable from the mail client it arrived in.

What it costs

Signing up puts the organisation on Free. Nothing expires and no card is asked for.

Free plan
monitors	3
min_interval_seconds	300
probe_locations	1
members	2
status_pages	1
history_window_days	90
agents	0
notifications_month	50
check_types	http
api_rate_limit	0

Read that table as: three monitors, probed no more often than every five minutes, from one location, with ninety days of history.

Three of those rows surprise people, so they are worth spelling out:

• check_types is http on Free. The other six types exist and are documented under Monitor types, but creating one on Free is refused with 403 feature_gated.
• api_rate_limit is 0 on Free, and that means the API is off — not throttled. Every /v1 request from a Free organisation is refused with 403 feature_gated, and so is creating an API key. The whole API reference applies to paid plans only; everything on this page is done in the dashboard.
• notifications_month is measured, never enforced. It is metered and shown in the dashboard, and passing it drops no alert and turns no usage bar red. Alerting is not something WireWatch will silently stop doing to you.

Where to go next

• Monitor types — the seven types and what each one checks.
• Check intervals — the allowed intervals, and why your plan may not offer the one you want.
• Probe locations — the regions, and running an agent inside your own network.
• Assertions — turning a reachable target into a correct one.
• API reference — driving all of the above from code, on a paid plan.