> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fluidify.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Routing

# Alert Routing Rules

Routing rules control what Regen does with an alert after deduplication and grouping. Rules are evaluated in priority order — the first matching rule wins.

Configure under **Settings → Routing Rules**.

***

## Match Criteria

The `match_criteria` object specifies which alerts a rule applies to. All fields are optional; omitting a field means "match anything". All specified fields must match (AND logic).

```json theme={null}
{
  "source":      ["prometheus", "grafana"],
  "severity":    ["critical", "warning"],
  "labels": {
    "env":       "prod",
    "alertname": "DiskUsage.*",
    "svc":       "*"
  },
  "annotations": {
    "summary": ".*connection refused.*"
  },
  "title":       ".*disk.*",
  "description": ".*OOM.*"
}
```

### Fields

| Field         | Type       | Matching                                              |
| ------------- | ---------- | ----------------------------------------------------- |
| `source`      | `string[]` | Exact list — alert source must be one of the values   |
| `severity`    | `string[]` | Exact list — alert severity must be one of the values |
| `labels`      | `object`   | Per-key match against alert labels (see below)        |
| `annotations` | `object`   | Per-key match against alert annotations (see below)   |
| `title`       | `string`   | RE2 regex matched against the alert title             |
| `description` | `string`   | RE2 regex matched against the alert description       |

### Label and annotation value matching

Each value in `labels` and `annotations` is matched independently:

| Value       | Meaning                                        |
| ----------- | ---------------------------------------------- |
| `"prod"`    | Exact string match                             |
| `"prod-.*"` | RE2 regex (matches `prod-eu`, `prod-us`, etc.) |
| `"*"`       | Key present with any non-empty value           |

Regex uses Go's [RE2 syntax](https://pkg.go.dev/regexp/syntax) — no lookaheads or backreferences.

### How `title` and `description` are derived

| Source     | `title`            | `description`         |
| ---------- | ------------------ | --------------------- |
| Prometheus | `labels.alertname` | `annotations.summary` |
| Grafana    | Rule name          | `annotations.summary` |
| CloudWatch | Alarm name         | Alarm description     |
| Generic    | `title` field      | `description` field   |

***

## Actions

| Action            | Key                               | Description                                          |
| ----------------- | --------------------------------- | ---------------------------------------------------- |
| Create incident   | `"create_incident": true`         | Automatically open an incident (default)             |
| Suppress          | `"suppress": true`                | Store the alert but create no incident               |
| Override severity | `"severity_override": "critical"` | Change the alert's severity before incident creation |
| Override channel  | `"channel_override": "db-oncall"` | Use a different Slack/Teams channel name suffix      |

```json theme={null}
{
  "create_incident":   true,
  "severity_override": "critical",
  "channel_override":  "db-oncall"
}
```

***

## Examples

### Suppress noisy disk-space alerts in staging

Drop alerts from the `staging` namespace that match a disk usage pattern — they fire constantly during CI and are not worth paging anyone.

```json theme={null}
{
  "match_criteria": {
    "labels": {
      "namespace": "staging",
      "alertname": "DiskUsage.*"
    }
  },
  "actions": {
    "suppress": true
  }
}
```

### Escalate any production database error to critical

Alerts from the `payments-db` service are automatically upgraded to `critical` severity regardless of what Prometheus reports.

```json theme={null}
{
  "match_criteria": {
    "labels": {
      "service": "payments-db",
      "env":     "prod"
    }
  },
  "actions": {
    "create_incident":   true,
    "severity_override": "critical",
    "channel_override":  "db-oncall"
  }
}
```

### Suppress alerts whose annotation mentions a known maintenance window

Use an annotation regex to catch alerts that already self-describe as expected.

```json theme={null}
{
  "match_criteria": {
    "annotations": {
      "summary": ".*scheduled maintenance.*"
    }
  },
  "actions": {
    "suppress": true
  }
}
```

### Catch-all: create an incident for everything else

Set a low-priority rule (high priority number) with an empty `match_criteria` as a fallback.

```json theme={null}
{
  "match_criteria": {},
  "actions": {
    "create_incident": true
  }
}
```

***

## Evaluation order

Rules are sorted by `priority` (ascending — lower number = higher priority). The **first** rule whose `match_criteria` matches the alert determines the action. Subsequent rules are not evaluated.

A catch-all rule (`match_criteria: {}`) at the end of the list acts as a default.