Docs / Alerts and incidentsBrowse

Alerts and channels

When a site fails 2 checks in a row, an incident opens and alerts go out. This page covers where they can go and how to choose who hears about which site.

On this page
  1. Email
  2. Phone push
  3. Slack, Microsoft Teams, Discord, Telegram and webhooks
  4. Webhook payload and signature
  5. SMS and phone calls
  6. Choose who hears about each site

Email

Plan/Every plan

Email is on for every new site. The owner, the admins and the site admins assigned to the site get one email when it goes down and one when it is back. The down email has an Acknowledge link and opens the incident. Times are in the workspace’s time zone.

When several sites go down at once, the first down email goes out right away. The ones that follow within a minute come together a minute later, as one email that lists the sites, each with its own Acknowledge link (or who already acknowledged it). A site that is back before then is listed as down and back up, and gets no separate email when it is back; every other site still does.

To add people, open the site, go to Settings and list up to 10 addresses under Also send to. They do not need an account.

Phone push

Plan/Every plan

Install the Heedery app, sign in, and press Turn on push alerts in its settings. Down alerts are marked time-sensitive, so they come through Focus. You can acknowledge from the lock screen. See Mobile app.

On the web, Settings then Notifications lists the phones signed in as you, sends a test push and removes a phone. Under What pushes it has a switch per kind of alert; the app has the same switches under Settings then Notifications. They are yours alone and hold for every phone you signed in on. Everything starts on.

On-call pages (important)
An escalation policy pages you: you are on call or named in a step. Its reminders and Escalate now too.
Site down
A site you get alerts for goes down and has no escalation policy.
Back up
A site that was down answers again, or someone resolves the incident.
Acknowledged by someone else
Someone else acknowledges an incident you were paged for or get alerts about: who, and when.
Checker blocked
A bot wall stops our checker, and when the site lets it through again.
Certificates and domains
A certificate or domain is about to expire, or was renewed.
Watched changes
A watched DNS record, certificate, page content or PageSpeed score changed, at the level set for the site.
Slowdowns
A site answers slower than usual, and when it is back to normal.
Site health
A new site health problem, like noindex, a missing analytics tag, a vulnerable WordPress or a leaked key, and when it is fixed.

A switch only stops the push. Email, channels, SMS and calls go out as before. Acknowledged by someone else is a push only: when someone takes an incident you were paged for or get alerts about, it says who and when, from the web, the app, the API or an email or channel link. You never get it for your own. See Escalation and on-call.

NoteOn-call pages is the push meant to wake whoever is on call, so turning it off asks first. A step that pages you then still emails you, and texts or calls you if you turned that on; the incident lists your push as skipped, with the reason. To be left alone off shift, use an on-call schedule instead.

The test push always goes out. To silence one phone only, turn off Heedery’s notifications in its system settings.

Slack, Microsoft Teams, Discord, Telegram and webhooks

Plan/Pro and up

  1. 1.
    Go to Settings then Notifications and pick a card under Channels.
  2. 2.
    Paste the address. How to set up in the dialog shows where to get it. Telegram needs no address: press Open in Telegram, or Add to a group, and Start the bot.
  3. 3.
    Choose when to Send the down alert and which Sites it covers: all sites, some categories, or selected sites.
  4. 4.
    Press Send test, then Add channel.
Settings, Notifications with the email panel and the Slack, Teams and webhook channels
Settings → Notifications: email for everyone, and the channels you add.
Slack
Uses an incoming webhook from your Slack app (api.slack.com/apps → Incoming Webhooks). Down messages have Acknowledge and Open in Heedery buttons.
Microsoft Teams
Uses a Teams Workflow: “Post to a chat or channel when a webhook request is received”. Older connector URLs work too.
Discord
Uses a channel webhook: channel settings → Integrations → Webhooks → New Webhook → Copy Webhook URL. Alerts are embeds in the status colour with Acknowledge and Open in Heedery links; they never mention anyone.
Telegram
Posts through the Heedery bot. The link in the dialog opens a private chat with it, or lets you pick a group; the bot says it is connected and the dialog shows the chat. Messages have Acknowledge and Open in Heedery buttons.
Webhook
Signed JSON to your own endpoint, on a public address. See below.

A channel can post the down alert immediately or wait from 1 to 60 minutes. An outage shorter than the wait never posts, and the recovery only follows a down alert that was sent. A workspace can have 30 channels.

If a Discord webhook is deleted, or the Heedery bot is blocked or removed from the Telegram chat, the channel shows Disconnected and nothing is sent there. Paste a new webhook URL, or connect the chat again, to resume. Discord and Telegram also get change, expiry, slowdown and blocked-checker warnings, like Slack.

When a send fails, Send test, the channel list and the incident timeline say in plain words what went wrong and what to do (a Slack webhook whose workspace is gone, a deleted Discord webhook, a Teams workflow that was turned off). What the service itself answered is kept under Details.

Webhook payload and signature

Each alert is a POST with a JSON body. The event is incident.down, incident.up or test.

incident.down
{
  "event": "incident.down",
  "site": { "id": "cm3…", "name": "Checkout API", "url": "https://api.example.com/health" },
  "incident": {
    "id": "cm4…",
    "startedAt": "2026-11-10T09:14:02.000Z",
    "endedAt": null,
    "durationSeconds": null,
    "error": "HTTP 503 Service Unavailable"
  },
  "workspace": { "id": "org_…", "name": "Acme" },
  "url": "https://heedery.com/admin/incidents/cm4…",
  "sentAt": "2026-11-10T09:15:03.000Z",
  "acknowledgeUrl": "https://heedery.com/ack/…"
}

url is the incident’s page in Heedery, in the down and the recovery alert alike, and site.url is the address being checked. The Open in Heedery button of Slack, Teams, Discord and Telegram messages opens the same incident page. Before 12 November 2026 the recovery alert’s url was the site’s page; to find the site, read site.id rather than the link.

Every request carries X-Heedery-Event, X-Heedery-Delivery and X-Heedery-Signature. The signature is sha256= followed by the hex HMAC-SHA256 of the raw body, keyed with the channel’s signing secret. The secret starts with whsec_ and is shown once, after you save the channel.

A failed send is retried with a growing pause, up to 5 times, for network errors, 408, 429 and 5xx answers. Other 4xx answers are not retried. Webhooks also receive change, expiry and slowdown warnings for the sites they cover.

SMS and phone calls

Plan/Pro and up. Through escalation policies only

Texts and calls go to people an escalation policy pages. Each person turns them on for themselves:

  1. 1.
    Go to Settings then Profile and find Phone alerts.
  2. 2.
    Enter your mobile number and press Send code. Only US and Canadian numbers work today.
  3. 3.
    Type the 6-digit code and press Confirm.
  4. 4.
    Under Alerts by phone, pick SMS, Call, or SMS and call.

A call reads the alert out twice. The all-clear comes by SMS only. An SMS uses 1 credit, a call 2. A text the carrier does not deliver, or a call nobody answers, gives its credits back.

Team includes 50 credits a month, Agency 200 and Enterprise 500. They are used first, renew each month and do not carry over. On top of them, the owner buys credits under Settings then Plan & billing: 100 for $10, 300 for $25, 1000 for $70. Purchased credits do not expire. With none left, texts and calls are skipped; email and push still go out.

Choose who hears about each site

Open the site, go to Settings, and look at the Alerts card:

When to alert
Wait a few minutes before opening the incident, so short blips never alert anyone.
Email alerts
Switch the email off for this site, or add more addresses.
Chat channels
Which channels cover this site. A channel set to Selected sites gets a switch here.
Escalation
A policy for this site, or the workspace default. See Escalation and on-call.

NoteA site with an escalation policy is paged by that policy only. Its steps replace the email and channel alerts above, and channel waits do not apply.

Still stuck? Write to us.