feat: temporal trends CLI script + Dash tab (Task D.1)

D.1a: cli/compute_trends.py — standalone CLI that imports existing pipeline
functions to replay pathway computation for ~10 historical 6-month endpoints.
Creates pathway_trends table via CREATE TABLE IF NOT EXISTS.

D.1b: Trends tab (10th PP tab) with metric toggle (patients/cost/cost_pp_pa).
Query gracefully returns empty when table doesn't exist, figure shows
instruction message to run compute_trends.

dash_app/data/queries.py

@@ -30,6 +30,7 @@ from data_processing.pathway_queries import (
     get_drug_network as _get_drug_network,
     get_drug_timeline as _get_drug_timeline,
     get_dosing_distribution as _get_dosing_distribution,
+    get_trend_data as _get_trend_data,
 )
 
 DB_PATH = Path(__file__).resolve().parents[2] / "data" / "pathways.db"
@@ -249,3 +250,12 @@ def get_dosing_distribution(
 ) -> list[dict]:
     """Average administered dose counts per drug."""
     return _get_dosing_distribution(DB_PATH, date_filter_id, chart_type, directory, trust)
+
+
+def get_trend_data(
+    metric: str = "patients",
+    directory: Optional[str] = None,
+    drug: Optional[str] = None,
+) -> list[dict]:
+    """Time-series trend data from pathway_trends table."""
+    return _get_trend_data(DB_PATH, metric, directory, drug)