# Decide on an alert

Handle an alert from start to finish, leaving a trace of your decision.

## The steps

1. Open **Risks**, under **Monitoring**.
2. Filter by **Severity**, **Category** or **Domain**, to start with what matters.
3. Read a row: what was observed, against what is expected. Expand **Actions to take** to see the recommended actions.
4. Click **Handle**, then choose a decision.
5. Add a comment if your team needs the context: a ticket number, the reason for an accepted risk.

## The decisions

| Decision | What it says | Where the alert goes |
|---|---|---|
| **To treat** | Someone is on it | It leaves the active alerts, for **In progress** |
| **Monitoring** | You keep an eye on it, without acting for now | **In progress** too |
| **Acknowledge** | Seen, and accepted | **Acknowledged** |
| **Reopen** | You cancel your decision | Back to the active alerts |

Each decision keeps its author and date. To set several at once, tick the rows: a bar appears with the same choices.

## Set an asset aside, openly

An asset should not alert you: equipment supplied by its vendor, whose obsolescence is not under your control, for example. In the same **Handle** menu, choose to exclude its alerts of that category, or **Exclude from all alerts**, then:

1. give a **reason**, which is required;
2. add an **end date** if its alerts should come back by themselves.

Nothing is deleted, and the whole company sees the exclusion. It is lifted from **Settings › My account › Analysis exclusions**.

> Acknowledging closes **one** alert. Excluding removes **the asset** from the analysis, for as long as the exclusion runs.

## An alert closes by itself

Once the problem is fixed, nobody has to close the alert: it is re-evaluated at the next collection, and resolves if the cause is gone.

### The Act button does not appear on a row

It requires a connector open to actions. The [Act on your infrastructure](/en/act) path explains the three locks that open it. The decision remains possible.

### An alert I acknowledged has come back

If it was resolved and the problem then reappeared, it is a new alert. For an asset that is permanently out of scope, prefer an exclusion with a reason.

### I get too many alerts on the same topic

Thresholds are set per company, with exceptions per asset: see [Costs and thresholds](/en/administer/costs-and-thresholds).
