fix(synthesis): Calculate analysis_period from actual data range

Previously hardcoded "Last 12 months" which was misleading when data
spanned multiple years. Now calculates the actual period from the
earliest to latest review dates in the dataset.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

packages/reviewiq-pipeline/src/reviewiq_pipeline/stages/stage5_synthesize.py

@@ -407,12 +407,14 @@ class Stage5Synthesizer:
     async def _gather_context(self, job_id: str) -> dict[str, Any]:
         """Gather all context needed for report generation."""
 
-        # Overview stats
+        # Overview stats with date range
         overview = await self.pool.fetchrow("""
             SELECT
                 COUNT(DISTINCT r.review_id) as total_reviews,
                 ROUND(AVG(r.rating)::numeric, 2) as avg_rating,
-                COUNT(s.span_id) as total_spans
+                COUNT(s.span_id) as total_spans,
+                MIN(r.review_time) as earliest_review,
+                MAX(r.review_time) as latest_review
             FROM pipeline.reviews_enriched r
             LEFT JOIN pipeline.review_spans s ON s.review_id = r.review_id
             WHERE r.job_id = $1::uuid
@@ -577,11 +579,30 @@ class Stage5Synthesizer:
             LIMIT 10
         """, job_id)
 
+        # Calculate analysis period from actual date range
+        earliest = overview["earliest_review"]
+        latest = overview["latest_review"]
+        if earliest and latest:
+            from datetime import datetime
+            days_span = (latest - earliest).days
+            if days_span <= 31:
+                analysis_period = "Last month"
+            elif days_span <= 90:
+                analysis_period = "Last 3 months"
+            elif days_span <= 365:
+                analysis_period = "Last 12 months"
+            else:
+                years = days_span // 365
+                analysis_period = f"Last {years} years" if years > 1 else "Last year"
+        else:
+            analysis_period = "All available data"
+
         return {
             "business_name": business or "This business",
             "total_reviews": overview["total_reviews"] or 0,
             "avg_rating": float(overview["avg_rating"] or 0),
             "total_spans": overview["total_spans"] or 0,
+            "analysis_period": analysis_period,
             "sentiment": [dict(r) for r in sentiment],
             "top_issues": [dict(r) for r in top_issues],
             "top_strengths": [dict(r) for r in top_strengths],
@@ -869,7 +890,7 @@ class Stage5Synthesizer:
                 generated_at=datetime.utcnow().isoformat(),
                 review_count=context["total_reviews"],
                 insight_count=context["total_spans"],
-                analysis_period="Last 12 months",
+                analysis_period=context.get("analysis_period", "All available data"),
                 charts=chart_data,
             )
 
@@ -1022,7 +1043,7 @@ Generate a comprehensive analyst report based on this data."""
             generated_at=datetime.utcnow().isoformat(),
             review_count=ctx["total_reviews"],
             insight_count=ctx["total_spans"],
-            analysis_period="Last 12 months",
+            analysis_period=ctx.get("analysis_period", "All available data"),
             charts=chart_data,
         )