schedule-review

Invoke the schedule-review skill with a scope such as `all` or a path glob like `src/MyService/**`. It runs TODO collection, suggestion scanning, and a structured checklist covering correctness, SOLID, async patterns, security, and test coverage, then merges results into a ranked report.

The TODO review layer collects `// TODO`, `// FIXME`, `// HACK`, and `// NOTE` comments plus open checklist items in documentation. Each item is classified as High, Medium, or Low priority and normalised into an actionable findings list.

Yes. After the consolidated report, findings matching the severity filter (blocking by default) are presented for confirmation. Approved findings become GitHub issues titled `[Review] <summary>` with labels `review`, `automated`, and a severity label.

The TODO and suggestion layers are language-agnostic, but the code-review checklist contains .NET-specific items such as async/await patterns and `dotnet test`. For other ecosystems, those checklist items should be adapted to the project's actual stack.

Yes. Each layer can be run independently by invoking its underlying skill directly (todo-review, suggestion-review, or code-review). You can also pass a comma-separated subset of layers to schedule-review instead of the default `all`.