Optimize AI generation speed and add richer insight data

Speed optimizations:
- Add session.prewarm() in InsightsViewModel and ReportsViewModel init
  for 40% faster first-token latency
- Cap maximumResponseTokens on all 8 AI respond() calls (100-600 per use case)
- Add prompt brevity constraints ("1-2 sentences", "2 sentences")
- Reduce report batch concurrency from 4 to 2 to prevent device contention
- Pre-fetch health data once and share across all 3 insight periods

Richer insight data in MoodDataSummarizer:
- Tag-mood correlations: overall frequency + good day vs bad day tag breakdown
- Weather-mood correlations: avg mood by condition and temperature range
- Absence pattern detection: logging gap count with pre/post-gap mood averages
- Entry source breakdown: % of entries from App, Widget, Watch, Siri, etc.
- Update insight prompt to leverage tags, weather, and gap data when available

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Shared/Services/FoundationModelsInsightService.swift

@@ -84,7 +84,13 @@ class FoundationModelsInsightService: ObservableObject {
         }
     }
 
-    /// Creates a new session for each request to allow concurrent generation
+    /// Prewarm the language model to reduce first-generation latency
+    func prewarm() {
+        let session = LanguageModelSession(instructions: systemInstructions)
+        session.prewarm()
+    }
+
+    /// Creates a fresh session per request (sessions accumulate transcript, so reuse causes context overflow)
     private func createSession() -> LanguageModelSession {
         LanguageModelSession(instructions: systemInstructions)
     }
@@ -213,8 +219,7 @@ class FoundationModelsInsightService: ObservableObject {
             throw InsightGenerationError.modelUnavailable(reason: lastError?.localizedDescription ?? "Model not available")
         }
 
-        // Create a new session for this request to allow concurrent generation
-        let session = createSession()
+        let activeSession = createSession()
 
         // Filter valid entries
         let validEntries = entries.filter { ![.missing, .placeholder].contains($0.mood) }
@@ -231,9 +236,10 @@ class FoundationModelsInsightService: ObservableObject {
         let prompt = buildPrompt(from: summary, count: count)
 
         do {
-            let response = try await session.respond(
+            let response = try await activeSession.respond(
                 to: prompt,
-                generating: AIInsightsResponse.self
+                generating: AIInsightsResponse.self,
+                options: GenerationOptions(maximumResponseTokens: 600)
             )
 
             let insights = response.content.insights.map { $0.toInsight() }
@@ -263,7 +269,7 @@ class FoundationModelsInsightService: ObservableObject {
 
         \(dataSection)
 
-        Include: 1 pattern, 1 advice, 1 prediction, and other varied insights. Reference specific data points.
+        Include: 1 pattern, 1 advice, 1 prediction, and other varied insights. Reference specific data points. Keep each insight to 1-2 sentences. If theme tags are available, identify what good days and bad days have in common. If weather data is available, note weather-mood correlations. If logging gaps exist, comment on what happens around breaks in tracking.
         """
     }