Python asyncpg fails with burst requests on both Supabase poolers: prepared statement errors on Transaction Pooler, timeouts on Session Pooler

# Bug report - [x] I confirm this is a bug with Supabase, not with my own application. - [x] I confirm I have searched the [Docs](https://docs.supabase.com), GitHub [Discussions](https://github.com/supabase/supabase/discussions), and [Discord](https://discord.supabase.com). ## Describe the bug SQLAlchemy with asyncpg is still creating prepared statements despite statement_cache_size=0 being set in connect_args, causing failures with Supabase's pgbouncer-based poolers (Shared Transaction Pooler on port 6543). This is triggered when there's a burst of requests to my server. Anything above a burst of 100 requests, it fails. The error occurs because prepared statements like "__asyncpg_stmt_9__" are being created and pgbouncer in transaction/statement mode doesn't support them, resulting in errors: prepared statement "__asyncpg_stmt_X__" does not exist. When I use the session pooler, the error because im `TimeoutError`. I've already upgraded from micro to small compute. Regardless if I use psycopg directly or sqlalchemy, both issues persist. This has been a long standing issue with supabase everytime I use it with python. It doesn't happen with apps I build using drizzle and typescript. ## To Reproduce ```python from typing import AsyncGenerator from uuid import uuid4 from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine from sqlalchemy.pool import NullPool from constants import DATABASE_URL # Use asyncpg for PostgreSQL async support database_url = DATABASE_URL.replace("postgresql://", "postgresql+asyncpg://") if ":5432" in database_url: # Create an engine with support for prepared statements engine = create_async_engine( database_url, echo=False, pool_pre_ping=True, connect_args={ "prepared_statement_name_func": lambda: f"__asyncpg_{uuid4()}__", "server_settings": { "jit": "off", "statement_timeout": "60000", # 60 seconds }, …