Suggested test: copilot/report-incomplete-job-failure

## Motivation

github/gh-aw#59316 (fixes github/gh-aw#59274) — `report_incomplete` now calls `core.setFailed()` so the Actions job conclusion becomes `failure`, even when issue reporting is disabled or fails. Previously the job stayed green.

## Proposed test

- **Workflow file**: existing `test-copilot-report-incomplete.md` needs its expected outcome updated, and/or a new variant `test-copilot-report-incomplete-create-issue-disabled.md`
- **Trigger**: `workflow_dispatch`
- **Engine**: copilot
- **Safe output**: `report-incomplete` with `create-issue: false`
- **Variant**: standard

## Minimal test prompt sketch

Agent calls `report_incomplete(reason: ...)` with `safe-outputs.report-incomplete.create-issue: false`. E2E harness should assert the workflow run **conclusion is `failure`**, not just that the run completed, and should NOT require an issue to be created.

## New fixtures or secrets needed

None. Requires updating `e2e.sh` validation logic (currently `trigger_workflow_dispatch_and_await_completion` / `wait_for_workflow` for the `report-incomplete` case appears to only check the run completed, not its conclusion) to assert `conclusion == "failure"` for this specific test case.

## Notes

The existing `test-copilot-report-incomplete.md` test currently expects/tolerates a green run. After gh-aw#59316 lands, that existing test's pass/fail semantics change — this should be reconciled (either update the existing test's expected conclusion, or add a distinct `create-issue:false` variant) before the harness reports false negatives.

> Generated by [🔍 Suggest New E2E Tests](https://github.com/githubnext/gh-aw-test/actions/runs/34205516021) · copilot · auto · 59.9 AIC · ⌖ 10.9 AIC · ⊞ 8.8K · [◷](https://github.com/search?q=repo%3Agithubnext%2Fgh-aw-test+is%3Aissue+%22gh-aw-workflow-call-id%3A+githubnext%2Fgh-aw-test%2Fsuggest-new-e2e-tests%22&type=issues)

Start with test-copilot-report-incomplete.md and inspect e2e.sh, especially trigger_workflow_dispatch_and_await_completion and wait_for_workflow. Decide whether to update the existing test or add the create-issue:false variant, then make validation require a failure conclusion without requiring an issue and run the relevant E2E test.

Written by the indexing model from the issue text.