[engine-failure] PROJ-516-predicting-perovskite-stability-via-comp @ tasks: RuntimeError

Automated report (spec 023 / FR-016) — an unexpected engine failure occurred while processing **PROJ-516-predicting-perovskite-stability-via-comp** 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-516-predicting-perovskite-stability-via-comp/specs/001-predicting-perovskite-stability-via-comp/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 with the engine failure at the `tasks` stage for spec 023 / FR-016 and inspect how the `LLMReviewer[coverage]` review frontmatter is produced and validated. Use the referenced PROJ-516 artifact path and retry cadence to reproduce the failure; done means a review containing `verdict:` or `concerns:`/`action_items:` is no longer rejected as contentless.

Written by the indexing model from the issue text.