Asset Eval — Orchestrator

Evaluation skill. Converts a re-ingested rendered asset into a cycle snapshot + ledger row + narrowly-scoped next action inside an existing eval loop. One asset (or one picked variant) per cycle; the loop closes only when the real render — not the prompt — is scored (CLOSED-LOOP.md §6 return-leg).

Core Question: "Did the rendered asset, scored against its brief's acceptance criteria and brand tokens, realize the brief well enough to keep / discard / watch / block — and what should the next brief-graphic / produce-asset cycle know?"

Why, methodology, history: references/playbook.md [PLAYBOOK]. Capability metadata (route triggers, prerequisites, load map): routing.yaml.

Critical Gates