Send alerts to email, Discord, or Telegram
Decide what is worth interrupting someone for, and where it lands.
Nothing is sent until you configure a channel and choose which risk levels may use it. Both decisions are per project.
Set up a channel
| Channel | What you need | Where to set it |
|---|---|---|
| SMTP host, port, username, password, from address | Settings → Email, then recipients per project | |
| Discord | A channel webhook URL | Project → Edit → Notifications |
| Telegram | A bot token and the chat ID | Settings → Telegram, then chat ID per project |
Using Gmail? Turn on two-factor authentication for the Google account, then create an app password under Google Account → Security → App passwords, and use that sixteen-character value as the SMTP password. Your normal password will not work.
Choose what is worth an alert
Open Project → Edit → Notifications and pick the risk levels allowed to notify. A message is sent when the incident's risk level is in that list. There is no separate score threshold — the level is the whole gate.
Starting with HIGH and CRITICAL only is the usual advice. Widening the net later is easy; recovering people's trust after a week of noise is not.
Repeat protection
A cooldown — thirty minutes by default — stops the same incident notifying repeatedly while it is still happening. Every attempt, including the ones skipped by cooldown, is recorded under Notifications.
Check it works
- Open Settings → Email and send a test message to yourself.
- If it arrives, open Notifications and confirm the attempt was logged.
- If it does not, the log entry states the reason — usually authentication or the wrong port.