OptimizedNodeBridge waitForAvailableWorker leaks timers after timeout

[bbopen]

Problem

waitForAvailableWorker() uses a polling loop via setTimeout(checkWorker, 10) but does not cancel the loop when the overall queue timeout fires. If the promise rejects on timeout, the polling continues indefinitely, creating a CPU/timer leak.

Evidence

• src/runtime/optimized-node.ts: waitForAvailableWorker sets an overall timeout and then recursively schedules checkWorker without a cancellation flag.

Acceptance criteria

• Stop polling when the queue timeout fires (e.g., set a flag and check it before rescheduling).
• Ensure no timers remain after the promise resolves or rejects.
• Add a test that forces queue timeout and verifies polling stops.

[bbopen]

🔄 Duplicate of #138

This issue describes the same timer leak in waitForAvailableWorker. The fix is tracked in #138 which was created as a follow-up from the PR #136 code review.