| name | alerting |
| description | Use when creating structured incident alerts from SysGuard findings, health reports, log analysis, remediation failures, or operational events. |
Alerting
Use this skill to turn operational findings into structured alerts that downstream notification or audit systems can consume.
Runtime Invocation
The Go implementation is registered by RegisterCoreSkills and invoked as "alerting":
registry.Execute(ctx, "alerting", &skills.SkillInput{Params: map[string]interface{}{
"severity": "critical",
"title": "Service down",
"message": "nginx is inactive",
"source": "sysguard",
}})
The result is an Alert with ID, severity, title, message, source, metadata, and timestamp.
Workflow
- Use
critical for down services, unreachable networks, or failed remediation.
- Use
warning for degraded CPU, memory, disk, or suspicious logs.
- Include source and metadata so later remediation can trace the alert.
- Pair with
notification when the alert needs to leave the process.
Safety
Alerts should describe facts observed by SysGuard. Do not invent root causes when the evidence only supports symptoms.