We check your APIs around the clock and tell you the moment one fails, by email and on your phone, and keep reminding you until it's fixed.
| Monitor | Started | Resolved | Duration | Reason |
|---|
| Sent | Type | Subject | To | Result |
|---|
Open a project to see its monitors. Members get alerts from the projects they're in.
People who asked for an account on the sign-in page. Approve them as a team admin (they get their own team), add them to an existing team, or reject them.
| Name | Role | Team | Monitors | Email recipients | Actions |
|---|
Your account is active, but you're not in a team yet. A team admin adds you to their team, and then your projects and monitors show up here.
Already added? Reload the page.
App Monitor calls your API URLs on a schedule. When one stops working it opens an incident and alerts you by email and phone, and when it works again it tells you it's back.
You can add monitors in projects where you're an owner or editor. In projects where you're a viewer, you can open a monitor to see how it's checked, but not change it.
On the Monitors tab, press + Add monitor.
Enter a Name and the URL. Pick the Method. Headers (such as an API key) and a request body go under Headers and body.
Check every 10 seconds to 24 hours. Timeout is how long to wait for an answer. Expected status is the HTTP code that means healthy.
200200-299200,204200-399Press Run test to call the API once and see its status, time and every field it returned. When it looks right, press Save. The first check runs straight away.
Up when the HTTP status matches Expected status. The body is never downloaded.
Use for APIs that return no body, such as pings, webhooks and HEAD requests. Run test offers this when the response is empty.
The status must match and every response check must pass.
"db": {"connected": false} ✗ downUse when an API can answer 200 while something behind it is broken, such as the database or a queue.
A check is field + condition + value. Fields use dots for objects and [0] for list items. Leave the field empty to check the raw response text, for HTML or plain-text answers. The quickest way to add a check: press Run test, then Add check next to the field you care about.
| Field | Condition | Value | Passes when |
|---|---|---|---|
status | equals | ok | status is exactly “ok” |
data.db.connected | equals | true | the database is connected |
queues[0].depth | is less than | 100 | the first queue has under 100 jobs |
version | exists | the field is there and not null | |
version | matches | ^2\. | a regular expression matches |
| (empty) | contains | OK | the raw text contains “OK” |
Also available: does not equal, does not contain and is greater than. All checks must pass. Field checks need a JSON response under 1 MB.
Every monitor of the picked project is one row. Down ones are always at the top, with a red stripe. The sample rows below are examples.
| Status | Monitor | Check | Alerts | Last checks | 24h |
|---|---|---|---|---|---|
| Down | Payments API Down 6m — HTTP 503, expected 200-299 ✋ I'm on itSnooze 1h |
200-299 +2 🔒 52d |
2× E 1h P each | 98.6% | |
| Maintenance | Worker queue GET https://…/status |
200 only |
1× E once P once | 100% |
Click 3 down, 95 up or 2 paused above the list to show only those (click again for all). Search by name or URL, and Sort by status, name, uptime, response time or last checked. Your choices are remembered in this browser.
Compact fits many monitors on one screen. Detailed shows each monitor as a card with its check and alert rules spelled out. Click any row or card to open the monitor.
The list refreshes every 5 seconds.
An incident starts at the first failed check, but is opened and emailed only once Alert after checks have failed in a row. Here Alert after is 3 failed checks. Every failure is also re-checked 5 seconds later before it counts, so a single dropped connection never alerts anyone.
By default you get one down alert per outage. To keep being reminded until it's fixed, open the monitor and set Repeat email while down and Repeat phone notification while down. Each has its own interval, for example phone On every failed check and email Every 30 minutes. Reminders arrive as “🔴 STILL DOWN: name (12m)” and stop as soon as the API recovers.
Down alerts have ✋ I'm on it and Snooze 1h: on the phone notification, in the email, and on the monitor in the list. I'm on it stops the reminders for everyone and tells the rest of the project who is handling it. Snooze pauses reminders for an hour. Both clear when the API recovers.
Deploying? In Projects → Manage → Maintenance, pause alerts for 30 minutes to 4 hours, or add a weekly window (e.g. Sundays 02:00–03:00). Checks keep running, but failures open no incident and send no alert; they show as blue bars.
Every https monitor's certificate is checked twice a day, with no setup. You're warned 14 and 3 days before it expires, and when it has expired. Days left show next to the check (🔒 52d).
The monitor's name, URL and method, the time it went down, the reason and how many checks failed in a row.
How long it was down, when it went down and recovered, and the HTTP status and response time of the check that passed.
The Incidents tab lists every outage, newest first. An incident still open shows Ongoing. Pausing or deleting a monitor closes its open incident.
App Monitor sends every email itself. You only choose who gets them, on the Settings tab.
One address per line, up to 20. Each person gets their own copy and can't see the others. New accounts start with their own address. If the list is empty, nothing is sent.
Tick Alert when a monitor goes down and Alert when it comes back up, or untick either one. They apply to email and phone notifications.
The last 24 hours of every monitor: uptime, downtime, incidents and average response time, sent at the time you pick, in your time zone (it starts as this browser's). Times in alert emails use the same time zone.
Press Send test email to check that emails arrive. Every email and phone notification sent, who it went to and whether it was delivered is listed on the Notifications tab. Change your password under Account on the same tab.
The Notifications tab lists every email and phone notification sent for the picked project, newest first: when, what, to whom, and whether it was delivered. Your own test emails and daily reports are there too.
Get down and up alerts (and the daily report) as notifications on your phone or computer, with nothing to install. Turn them on in Settings → Phone notifications.
Monitors live in projects. Your team admin adds you to projects and gives you a role in each. Pick a project in the Project picker at the top, and Monitors, Incidents and Notifications show that project. Your role in it shows next to the picker.
Monitors live in projects. You create projects on the Projects tab with + New project, then add people from your team in Manage and give each one a role. People only see the projects they're added to.
| Role | See monitors and incidents | Add, edit, delete monitors | Add or remove people |
|---|---|---|---|
| Owner | ✓ | ✓ | Editors and viewers; change roles; hand the project over |
| Editor | ✓ | ✓ | Viewers only |
| Viewer | ✓ | — | — |
One card per project: your role, how many monitors (and how many are down, in red) and members. Click a card to open it in Monitors. All in my team shows every project in your team, and Search finds one by name.The Project picker at the top lists the same projects.
Press Manage on a card for its name, My alerts, Maintenance windows, Members (roles, add, remove), and Delete or Leave.
Press Alerts on a card for My alerts and Leave project. Owners and editors see Manage, with maintenance windows and members too.
As a team admin you look after the people and projects of your team. You can't see other teams.
On the TeamUsers tab, enter their email, name, a first password (10+ characters) and the role: Member, or Team admin to share the work. Give them the email and password yourself.
On Projects, press + New project, then add people in Manage as Editor (can change monitors) or Viewer (read-only, chooses their own alerts).
Projects → All in my team shows every project in your team, even ones you aren't in, and you can manage them all.
On the TeamUsers tab, View as opens App Monitor as that person for up to an hour, to see exactly what they see and fix their settings. A yellow banner shows it, Back to my account ends it, and every use is logged.
Set password gives someone a new password and signs them out everywhere. Delete removes the account; projects they own pass to another member.
As superadmin you run App Monitor for everyone: accounts, teams and what each team may monitor.
People ask for an account with Request access on the sign-in page. You get a phone notification, and the request appears at the top of the Users tab. Approve them as admin of a new team or into an existing team, or Reject. Until then they can't sign in.
Give someone the Team admin role and they manage their own team. Someone made team admin without a team gets a new one.
Rename teams, move people between them, and delete empty ones. Only teams you tick May monitor private addresses can check internal services (10.x, 127.0.0.1, cluster names); everyone else can only monitor public URLs.
You don't open other teams' projects directly. Use View as on the Users tab to work as that person for up to an hour; it's logged. Your own projects stay private to team admins unless you add them.
| Reason shown | What it means | What to do |
|---|---|---|
HTTP 500, expected 200-299 | The API answered with a different status. | Fix the API, or change Expected status if that code is normal (for example 204). |
Timed out after 10s | No answer within the timeout. | Check the API is running. Raise Timeout if it is just slow. |
ENOTFOUND (host) | The host name doesn't resolve. | Check the URL for typos and that the DNS record exists. |
ECONNREFUSED | Nothing is listening on that address and port. | Check the service is up and the port is right. |
Check failed: … got … | A response check didn't pass. It shows what came back. | Open the monitor, press Run test and compare the field with your check. |
points to a private address | The URL is on a private or internal network (10.x, 127.0.0.1, cluster names), which your team isn't allowed to monitor. | Use the service's public URL. Ask your superadmin if your team really needs internal addresses.Tick May monitor private addresses for the team on the Users tab. |
This alert link has expired | The I'm on it / Snooze links in an alert email work for 2 days. | Open App Monitor and use the buttons on the monitor instead. |
Response is not valid JSON | A field check ran on a non-JSON answer. | Use an empty field to check the raw text, or switch to Status code only. |
Good to know: you stay signed in in this browser tab for 7 days from your last visit. After 10 wrong passwords, sign-in is blocked for 15 minutes. Check history is kept 8 days, the email log 30 days, and incidents until the monitor is deleted.