Supabase query performance

Description

Diagnose and fix Postgres performance in Supabase projects — finding slow queries (Query Performance report, pg_stat_statements, Performance Advisor, index_advisor), reading EXPLAIN (ANALYZE, BUFFERS) plans, choosing indexes (B-tree composite order, partial, covering, GIN for JSONB/arrays/full-text, trigram for ILIKE, BRIN for append-only time series), making RLS policies fast, connection management (Supavisor transaction vs session pooling, prepared statements, serverless clients), pagination and counts in supabase-js/PostgREST, N+1 and embedded selects, batch writes and upserts, lock contention and short transactions, statement timeouts, vacuum/analyze and bloat, Realtime scale, and partitioning. Use whenever a Supabase/Postgres query or page is slow, the database CPU or connections are high, the user sees timeouts or "too many connections", reviews query/index changes, or prepares for scale.