[engine-failure] PROJ-884-llmxive-follow-up-extending-self-improvi @ tasks: RuntimeError

Automated report (spec 023 / FR-016) — an unexpected engine failure occurred while processing **PROJ-884-llmxive-follow-up-extending-self-improvi** at stage `tasks`.

```
RuntimeError: LLMReviewer[coverage]: review frontmatter has neither `verdict:` nor `concerns:`/`action_items:` — a contentless/truncated review MUST NOT count as acceptance (false-convergence guard). First 200 chars: '```yaml\n---\nreviewer_name: coverage\nreviewer_kind: llm\nstage: tasked\nartifact_path: projects/PROJ-884-llmxive-follow-up-extending-self-improvi/specs/001-llmxive-follow-up-extending-self-improvi/tasks.'
```

The project remains schedulable; it will retry on its normal cadence and recover once the underlying defect is fixed. Closing this issue does not require any project-state change.

Start by tracing the LLMReviewer[coverage] handling for the `tasks` stage and inspect the referenced artifact under `projects/PROJ-884-llmxive-follow-up-extending-self-improvi/specs/001-llmxive-follow-up-extending-self-improvi/tasks`. Reproduce the retry failure with the truncated review shown in the report; done means a review lacking `verdict:` and `concerns:`/`action_items:` cannot count as acceptance and the engine no longer fails unexpectedly.

Written by the indexing model from the issue text.