Add BeanFlows MVP: coffee analytics dashboard, API, and web app

- Fix pipeline granularity: add market_year to cleaned/serving SQL models
- Add DuckDB data access layer with async query functions (analytics.py)
- Build Chart.js dashboard: supply/demand, STU ratio, top producers, YoY table
- Add country comparison page with multi-select picker
- Replace items CRUD with read-only commodity API (list, metrics, countries, CSV)
- Configure BeanFlows plan tiers (Free/Starter/Pro) with feature gating
- Rewrite public pages for coffee market intelligence positioning
- Remove boilerplate items schema, update health check for DuckDB
- Add test suite: 139 tests passing (dashboard, API, billing)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

transform/sqlmesh_materia/models/cleaned/cln_psdalldata__commodity_pivoted.sql

@@ -13,6 +13,7 @@ SELECT
   max(commodity_name) as commodity_name,
   country_code,
   max(country_name) as country_name,
+  market_year,
   ingest_date,
   COALESCE(SUM(CASE WHEN attribute_name = 'Production' THEN value END), 0) AS Production,
   COALESCE(SUM(CASE WHEN attribute_name = 'Imports' THEN value END), 0) AS Imports,
@@ -48,8 +49,10 @@ WHERE attribute_name IN (
 GROUP BY
   commodity_code,
   country_code,
+  market_year,
   ingest_date
 ORDER BY
   commodity_code,
   country_code,
+  market_year,
   ingest_date