fix: fail reporter worker when final result flush cannot send results

[dtice25]

Summary

BenchmarkReporter.flush catches sendTaskResults errors and leaves unsent records in this.pending, but finish still called completeWorker if flush(true) did not throw. This allowed a worker (and therefore the run) to be marked complete before all of its task results reached the platform, which could also make a run appear complete sooner than it actually was.

Changes:

• Track flushFailed on the reporter and set it when any batch fails to send.
• finish now checks flushFailed and pending.length after the final flush. If any result remains unsent, it calls failWorker instead of completeWorker.
• Only call completeWorker when the final flush has accepted every pending result.

if (this.flushFailed || this.pending.length > 0) {
  await this.client.failWorker(..., new Error('Failed to flush task results'));
  return;
}
await this.client.completeWorker(...);

Prevents a worker from reporting success when its result batch could not be delivered.

Link to Devin session: https://app.devin.ai/sessions/06813b4b9fb64b60b1f47bd8e1e8f152
Requested by: @dtice25

---

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring

[open-cla]

Contributor License Agreement

All contributors are covered by a CLA.

[devin-ai-integration]

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.