fix(queue): harden drain/abort/timeout race handling

- reject new lane enqueues once gateway drain begins
- always reset lane draining state and isolate onWait callback failures
- persist per-session abort cutoff and skip stale queued messages
- avoid false 600s agentTurn timeout in isolated cron jobs

Fixes #27407
Fixes #27332
Fixes #27427

Co-authored-by: Kevin Shenghui <shenghuikevin@github.com>
Co-authored-by: zjmy <zhangjunmengyang@gmail.com>
Co-authored-by: suko <miha.sukic@gmail.com>

src/config/sessions/types.ts

@@ -84,6 +84,14 @@ export type SessionEntry = {
   spawnDepth?: number;
   systemSent?: boolean;
   abortedLastRun?: boolean;
+  /**
+   * Session-level stop cutoff captured when /stop is received.
+   * Messages at/before this boundary are skipped to avoid replaying
+   * queued pre-stop backlog.
+   */
+  abortCutoffMessageSid?: string;
+  /** Epoch ms cutoff paired with abortCutoffMessageSid when available. */
+  abortCutoffTimestamp?: number;
   chatType?: SessionChatType;
   thinkingLevel?: string;
   verboseLevel?: string;