chore(cli-repl): try to fix flaky connection string prompt test

[addaleax]

Currently, in CI we receive errors like

1) CLI entry point
    asks for connection string when configured to do so:
  Uncaught Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:161:15)
    at writeGeneric (node:internal/stream_base_commons:152:3)
    at Socket._writeGeneric (node:net:958:11)
    at Socket._write (node:net:970:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Socket.Writable.write (node:internal/streams/writable:510:10)
    at Socket.<anonymous> (Z:\data\mci\e6e6\src\packages\cli-repl\src\run.spec.ts:87:20)
    at Socket.emit (node:events:524:28)
    at Socket.emit (node:domain:489:12)
    at addChunk (node:internal/streams/readable:561:12)
    at readableAddChunkPushByteMode (node:internal/streams/readable:512:3)
    at Socket.Readable.push (node:internal/streams/readable:392:5)
    at Pipe.onStreamRead (node:internal/stream_base_commons:191:23)
    at Pipe.callbackTrampoline (node:internal/async_hooks:130:17)

(e.g. https://spruce.mongodb.com/task/mongosh_tests_win32_m50xc_n20_test_cli_repl_patch_7fa888fae7921991c99175aad0ebba46751c2954_68907a2301ef2900071ba30e_25_08_04_09_15_17/logs).

The hypothesis here is that the following sequence of events occurs:

1. The test runs normally up to the point where the application prompts for Press any key to exit.
2. The test reads that text from the application and sends and x key to the mongosh process.
3. The mongosh process reads the x key and echoes it back to the terminal (since it is in raw mode).
4. The mongosh process exits (the test passes if this happens after the next steps, which is where the flakiness of the test may come from).
5. The test reads the x key echoed back in step 3 and, seeing more output coming from mongosh, repeats step 2, since its condition is still met.
6. Sending the second x key to the process fails with EPIPE because the target process has already exited.

Sending the x key only once should resolve the flakiness, if this is indeed what happens.

[Copilot]

Pull Request Overview

This PR addresses a flaky test in the CLI REPL component by preventing duplicate key writes to an already exited process. The issue was that the test would sometimes send an 'x' key twice to the mongosh process when prompted to "Press any key to exit", causing the second write to fail with EPIPE after the process had already terminated.

• Introduces a flag to track whether the "any key to exit" response has been sent
• Modifies the condition to only send the 'x' key once when the exit prompt appears

Comments suppressed due to low confidence (1)

packages/cli-repl/src/run.spec.ts:77

• [nitpick] The variable name wroteAnyKeyToExit is inconsistent with the existing pattern. Consider renaming to wroteAnyKey to match the pattern of wroteConnectionString.

    let wroteAnyKeyToExit = false;

[gagik]

Nice, a very subtle thing. Hopefully it fully solves this.