Persist GSC comparable-window delta events and expose latest alerts retrieval

[AJaySi]

Motivation

• Avoid repeated heavy per-page/query comparisons in the UI by precomputing and persisting GSC delta events for dashboard and content-strategy flows.
• Provide a single, indexed storage surface that other services can query for recent opportunities/alerts instead of recalculating deltas on demand.
• Support actionable classifications (decline/rise/opportunity) and severity metadata so downstream features can prioritize signals.

Description

• Added PlatformInsightDeltaEvent model/table to persist delta events with ownership, event_type, entity_type, entity_key, comparable-window metadata, details payload, severity, and relevant indexes (backend/models/platform_insights_monitoring_models.py).
• Enhanced GSCInsightsExecutor to fetch the current 30-day window and the prior comparable 30-day window, compute per-page and per-query deltas (clicks/impressions/CTR/position), label events using in-code thresholds (decline/rise/opportunity), and persist events to platform_insight_delta_events while deduplicating by current window (backend/services/scheduler/executors/gsc_insights_executor.py).
• Added OpportunityAlertsService to provide a lightweight, user-scoped retrieval API for the latest persisted events with filtering by platform/site/event types (backend/services/analytics/opportunity_alerts_service.py).
• Exposed get_latest_opportunities_alerts(...) on PlatformAnalyticsService to let UIs and analytics layers fetch precomputed events without recomputing deltas (backend/services/analytics/platform_analytics_service.py).
• Persistence includes deterministic dedupe (delete previous rows for the same user/platform/site/current-window) and stores a full details JSON containing metrics, deltas, reasons, thresholds, and window metadata.

Testing

• Compiled modified Python modules with python -m py_compile backend/models/platform_insights_monitoring_models.py backend/services/scheduler/executors/gsc_insights_executor.py backend/services/analytics/opportunity_alerts_service.py backend/services/analytics/platform_analytics_service.py, which succeeded without syntax errors.
• Performed lightweight static validation by importing/compiling the changed modules (no runtime integration tests executed in this rollout).

---

Codex Task