Alert time-based and conditional snoozing

**What the feature is (as Title)**
Alert time-based and conditional snoozing

**Value proposition**

When an alert fires during an active investigation, teams often face a bad choice: keep getting paged for the same condition, or disable the rule and risk missing the moment things get worse. Without snooze, every repeat notification adds noise, slows triage, and makes it harder to focus on resolving the incident.

This feature lets users define when an alert should be snoozed (no actions fire while snoozed), for how long, and under what conditions it should un-snooze. That gives operators flexibility to quiet known or in-progress issues without turning alerting off entirely, and a systematic way to restore notifications when the incident is actually over or when condition escalates and attention is needed again.

**Expected outcome**

Teams investigating an incident can pause unwanted notifications without abandoning coverage: on-call engineers stop getting repeat pages for alerts they are already working, while rules keep evaluating and history stays intact. When the situation changes, because time runs out, severity drops, or conditions no longer match, the right notifications come back on their own, so teams do not have to remember to re-enable rules or manually undo blanket silences. The result is less alert fatigue during triage, clearer handoffs (snoozes are deliberate and time-bound), and more confidence that paging will resume when the incident is truly concluded or when it escalates again.