fix(db): stop 'set sqlite pragmas' lock flake on concurrent connects

Set busy_timeout before the WAL switch and retry the pragma batch on
'database is locked' within the same 5 s budget: while a file is still in
rollback mode, a shared-lock holder upgrading for the journal_mode change
gets SQLITE_BUSY without the busy handler running, so busy_timeout alone
doesn't cover parallel opens of a fresh file. Adds a 16-thread stress test
and drops the per-module DB_INIT workaround in ai::handlers tests.

Closes #157

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

src/ai/handlers.rs

@@ -2129,15 +2129,7 @@ mod turn_replay_tests {
     use actix_web::{App, web::Data};
     use std::sync::Arc;
 
-    /// Serialize `AppState::test_state()` construction across the parallel
-    /// tests in this module: each build opens ~10 DAO connections to the one
-    /// shared `DATABASE_URL` file, and doing several at once races the WAL
-    /// `journal_mode` switch into a spurious "database is locked". The test
-    /// bodies themselves still run in parallel; only the open is gated.
-    static DB_INIT: std::sync::Mutex<()> = std::sync::Mutex::new(());
-
     fn build_state() -> Data<AppState> {
-        let _guard = DB_INIT.lock().unwrap_or_else(|p| p.into_inner());
         Data::new(AppState::test_state())
     }