← Back to Documentation
Continuous Monitoring — Guide
Get MITRE & framework-tagged security drift alerts in Microsoft Teams.
Continuous Monitoring keeps watch over your Microsoft 365 security posture and alerts you in Microsoft Teams the moment something drifts — a passing control starts failing, a policy weakens, or a new gap appears. Every alert is tagged with the MITRE ATT&CK technique and the compliance frameworks it affects. This is a separate product from the one-time assessment report.
What Continuous Monitoring does
- Detects configuration drift. Kyvoi re-checks your tenant on your chosen schedule and compares it to the last known state.
- Alerts only on change. No change means no message — no noise.
- Tags every alert with context — control, severity, MITRE ATT&CK technique, and compliance clauses impacted.
- Delivers into Microsoft Teams as rich, actionable cards with deep links to fix the issue in the Microsoft portal.
Monitoring complements — not replaces — Microsoft Defender. Defender detects active threats; Kyvoi detects posture drift and compliance regressions in your configuration.
How alerts are delivered
Alerts go to a Microsoft Teams channel using an incoming webhook you create and control. Kyvoi never connects into your network — it sends an outbound message to the webhook URL you provide. No inbound connections, no agents, and no firewall changes.
Noise control (three tiers)
- Instant alerts — Critical and High regressions are pushed immediately.
- Digest — Lower-severity changes are batched into a periodic summary.
- Silence — Nothing changed since the last check means nothing is sent.
Setting up your Teams webhook URL
- In Microsoft Teams, open (or create) the channel where you want alerts — for example, Kyvoi Alerts.
- Next to the channel name, click the ⋯ (More options) menu.
- Choose Workflows.
- Select "Post to a channel when a webhook request is received" (in some tenants: "Send webhook alerts to a channel").
- Confirm the Team and Channel, then create the workflow.
- Copy the webhook URL it generates — this is what you paste into Kyvoi.
Teams identity: when creating the workflow, sign in with an account that is a member of the target team, so the workflow is allowed to post to that channel.
Keep the URL private. Anyone with your webhook URL can post to your channel. If it's ever exposed, delete the workflow and create a new one.
Configuring monitoring in Kyvoi
On the monitoring settings page you provide three things:
- Microsoft Teams webhook URL — the URL you copied above.
- Instant-alert severities — which severities alert immediately (default: Critical and High). The rest are batched into a digest.
- Check frequency — every 15 minutes, 30 minutes, hour, 2 hours, or 24 hours.
More frequent checks mean faster detection. Billing is metered by check volume. Click Save, and a confirmation tells you the cadence and channel.
What each alert contains
- Control — the specific control that changed.
- Change — the status transition (e.g. PASS → FAIL).
- Severity — Critical, High, Medium, or Low.
- MITRE ATT&CK — the related technique (e.g. T1110 — Credential Access).
- Compliance impact — the framework clauses affected.
- Actions — deep links to fix in the Microsoft portal and view the MITRE technique.
Security and privacy
Agentless. Kyvoi installs nothing in your environment.
Read-only. Monitoring uses read-only Graph permissions. No settings are changed.
Outbound only. Kyvoi reads via Graph and sends alerts to your Teams webhook — both outbound to Microsoft. No inbound connection to your network.
You control delivery. Alerts go only to the webhook and channel you choose.
Frequently asked questions
Do I need to open a firewall port or allow-list Kyvoi?
No. There are no inbound connections to your network. All calls are outbound to Microsoft.
Will I get spammed with alerts?
No. Alerts fire only when your posture changes; lower-severity changes are batched into a digest.
Can I change the frequency or channel later?
Yes — re-open the settings and update the webhook, severities, or frequency anytime.
Is this the same as the one-time report?
No — the report is a snapshot; monitoring is an ongoing watch. They are separate products.