Add **Shared ScheduledExecutorService** for timeouts

[edburns]

CopilotSession.java

• Added ScheduledExecutorService import.
• New field timeoutScheduler: shared single-thread scheduler, daemon thread named sendAndWait-timeout.
• Initialized in 3-arg constructor.
• sendAndWait(): replaced per-call Executors.newSingleThreadScheduledExecutor() with timeoutScheduler.schedule(). Cleanup calls timeoutTask.cancel(false) instead of scheduler.shutdown().
• close(): added timeoutScheduler.shutdownNow() before the blocking session.destroy RPC call so stale timeouts cannot fire after close.

TimeoutEdgeCaseTest.java (new)

• testTimeoutDoesNotFireAfterSessionClose: proves close() cancels pending timeouts (future not completed by stale TimeoutException).
• testSendAndWaitReusesTimeoutThread: proves two sendAndWait calls share one scheduler thread instead of spawning two.
• Uses reflection to construct a hanging JsonRpcClient (blocking InputStream, sink OutputStream).

Resolves #ISSUE_NUMBER

---

[edburns]

01-mock-llm-to-show-blocking.md

Prompts that produced this PR.

[Copilot]

Pull request overview

This PR updates CopilotSession#sendAndWait to use a session-scoped ScheduledExecutorService instead of creating a new scheduler per call, and adds regression tests covering timeout behavior during close() and scheduler thread reuse.

Changes:

• Add a session-level timeoutScheduler and use it for sendAndWait() timeouts.
• Update close() to shut down the timeout scheduler before invoking the blocking session.destroy RPC.
• Add TimeoutEdgeCaseTest to verify timeouts don’t fire after close() and that timeout scheduling reuses a single thread.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.

File	Description
src/main/java/com/github/copilot/sdk/CopilotSession.java	Introduces a shared per-session scheduler for sendAndWait timeouts and shuts it down during close().
src/test/java/com/github/copilot/sdk/TimeoutEdgeCaseTest.java	Adds tests for timeout-after-close and for scheduler thread reuse across multiple sendAndWait calls.

Comments suppressed due to low confidence (2)

src/test/java/com/github/copilot/sdk/TimeoutEdgeCaseTest.java:125

• Second Thread.sleep(100) has the same flakiness risk as the first. Consider using a bounded wait/polling approach instead of fixed sleeps to make the test deterministic.

            Thread.sleep(100);
            long afterSecond = countTimeoutThreads();

src/test/java/com/github/copilot/sdk/TimeoutEdgeCaseTest.java:114

• If an assertion fails before reaching the explicit session.close() at the end of the test, the session/scheduler thread can leak. Consider try-with-resources for CopilotSession or a finally that always closes it.

            CopilotSession session = new CopilotSession("test-thread-count-id", rpc);

            long baselineCount = countTimeoutThreads();

            CompletableFuture<AssistantMessageEvent> result1 = session
                    .sendAndWait(new MessageOptions().setPrompt("hello1"), 30000);

[brunoborges]

@edburns let me know when you finish clearing the feedback from Copilot on this PR, then flag this PR as Ready for Review and I'll take a look!

[edburns]

@brunoborges , this is now ready for your review. I'd appreciate your input. This is preparatory work for allowing the API to accept an ExecutorService, as we briefly discussed last week.

[edburns]

Marking this as Draft in preparation for introducing the same PR but from a topic branch on this repository, rather than from a topic branch on a fork of this repository.

[edburns]

Is superseded by #38