loop-execution-evaluator

You route track evaluation by providing the track's metadata.json and spec.md to the dispatcher, which determines the track type and dispatches execution to the correct specialized evaluator for quality assessment.

The best way to apply correct quality gates is using a dispatcher that reads track metadata to route execution to specialized evaluators like eval-ui-ux, eval-code-quality, eval-integration, or eval-business-logic as appropriate.

The dispatcher aggregates evaluator results by collecting outputs from the dispatched specialized evaluators into a single final verdict for the Evaluate-Loop and then triggers subsequent steps in the workflow.

Yes, you need both metadata.json and spec.md files, as the dispatcher reads these inputs to determine the track type and route execution to the correct specialized evaluator workflow.

Yes, you can trigger subsequent workflow steps because the dispatcher aggregates results from dispatched evaluators into a single final verdict for the Evaluate-Loop and coordinates the following actions.

The specialized evaluators available for routing track evaluation are eval-ui-ux, eval-code-quality, eval-integration, and eval-business-logic, each targeting specific quality assessment domains based on track type.