loop-execution-evaluator

To coordinate automated evaluation across multiple review tracks, this skill dispatches tasks to specialized evaluators based on track type and aggregates their results into a single consolidated verdict.

Dispatching evaluators by track type works by reading the track metadata.json and spec.md files to determine the track category, then routing the evaluation to the appropriate specialized evaluator like eval-ui-ux or eval-code-quality.

You can use a single coordinator to dispatch to specialized evaluators for UI, code-quality, integration, and business-logic tracks, avoiding a one-size-fits-all checklist by matching each track to its correct evaluator.

The best way to aggregate results from different code review evaluators is using a coordinator that dispatches to specialized evaluators and produces a consolidated verdict, ensuring accurate track-specific assessment.

Yes, you need metadata.json and spec.md files to route track evaluations, as the dispatch logic reads these files to determine the track type and coordinate the appropriate specialized evaluators.