fix(alerts): Populate metric_alert webhook actions for native monitors

[Christinarlong]

Summary

When a Metric Monitor (workflow_engine Detector) fires and notifies a Sentry App, the metric_alert webhook payload had an empty data.metric_alert.alert_rule.triggers[].actions array — so the Sentry App action and its configured settings (e.g. a channel select) never reached the app.

The payload is rebuilt from workflow_engine models by WorkflowEngineDataConditionSerializer.get_attrs, which attaches actions to a trigger only when the workflow's action-filter DataConditionGroup contains an ISSUE_PRIORITY_GREATER_OR_EQUAL condition whose comparison matches the trigger's priority (HIGH=75 / MEDIUM=50):

if trigger.condition_result in dcg_comparison_pairs.get(dcg_id, set())

• Migrated alert rules always create that priority condition → match succeeds. ✅
• Natively-created Metric Monitor connected alerts seed the action filter with empty conditions (automationBuilderContext.tsx). With no priority condition, nothing matched and the actions were dropped → actions: []. ❌

Fix

Track whether each action-filter DCG gates on priority at all, and treat a DCG with no priority gate as firing for any priority — so its actions are attributed to the detector's triggers. Migrated rules still take the priority-match branch and are unchanged; only priority-less (native) filters take the new path.

A priority-less filter matches all of the detector's triggers, which is semantically correct (an empty filter fires for any priority); the webhook's top-level status still identifies which trigger fired.

Fixes #118042
Refs ENG-7988