Notifications
Configure how and where Cassian sends you alerts — email, Telegram, or webhook.
Cassian™ can alert you the moment something changes — a scan completes, your store goes down, or an order anomaly is detected. You decide which events matter to you, and where those alerts arrive.
Notification channels
Cassian supports three channels:
| Channel | What it is |
|---|---|
| Alerts and digests sent to every organisation admin | |
| Telegram | Real-time alerts via the Cassian Telegram bot |
| Webhook | HTTP POST to any HTTPS URL — integrates with Slack, Discord, or custom systems |
Channel availability by plan
Not all channels are available on every plan.
Email is available on every plan, including the free Inspector tier. Telegram and webhook notifications are available on the Protector plan and above.
What can trigger a notification
Cassian sends notifications from ten sources. Each source can be toggled per channel in the "What to send where" matrix:
- Scan complete — When a full site scan finishes, including your Cassian Score™ and a findings summary
- Uptime alert — When Cassian Sentinel™ detects your store is down (and when it recovers), plus SSL and domain expiry warnings
- Order Pulse incident — When an order anomaly is detected (and when orders return to normal)
- Security finding — When Cassian Shield™ finds new vulnerabilities
- Link check — When new broken internal links are found
- Pricing Pulse — When product prices change unexpectedly, including bulk price changes
- Stock Monitor — When a product goes out of stock, or is projected to run out soon
- Sitemap — When a large number of URLs disappear from your sitemap since the last check
- PageSpeed — When a sustained performance regression is confirmed (not one-off measurement blips)
- Week in review — A weekly digest: your score and its change, scans run, issues resolved, new issues, and open issue counts
The weekly digest's day and hour are configurable in Settings, and are interpreted in your organisation's timezone. If you haven't set a preference, it sends on Monday at 08:00 UTC.
Where to configure
All notification settings live in Settings → Notifications.
The What to send where section shows a matrix of event sources (rows) × channels (columns). Toggle each combination on or off to control exactly what you receive and where. Everything is on by default — you opt out of what you don't want.
Severity routing and quiet hours
Beyond the per-source matrix, Cassian gives you three finer controls (also in Settings):
- Per-channel severity floors — Set a minimum severity for each channel. For example, keep email at "info" (everything) but set Telegram to "critical" so your phone only buzzes when it matters. Severities are critical, warning, and info.
- Quiet hours — Define a nightly window in your organisation's timezone. Warning and info alerts raised during quiet hours are held until the window ends. Critical alerts always break through — if your store goes down at 3am, you'll hear about it.
- Per-site monitor sensitivity — Each monitor (Order Pulse, Pricing Pulse, Stock, Uptime, links, sitemap, PageSpeed) can be tuned per site, so a busy store and a quiet side project don't have to share thresholds.
Ongoing incidents live on the Alerts page (/dashboard/alerts) — one timeline for every monitor, from first detection through acknowledgement, snooze, and recovery. Admins can acknowledge or snooze incidents there, and the page also shows a log of recent notification deliveries so you can confirm what was actually sent. See Alerts for the full picture.
How suppression works
Cassian won't send you the same alert repeatedly for an ongoing issue. Each source has its own re-notification window — repeated alerts for the same logical event are rate-limited:
| Source | Minimum gap between repeat alerts |
|---|---|
| Uptime down / recovery | 5 minutes |
| Security findings | 1 hour |
| Pricing Pulse | 1 hour |
| Broken links, sitemap, PageSpeed, stock | 24 hours |
| SSL expiry warnings | 7 days |
| Order Pulse | Follows its own incident cadence (re-notifies based on incident type) |
If an alert keeps arriving for something you already know about, acknowledge or snooze the incident from the Alerts page rather than turning the whole source off.