Sakila execute-tier fixture for the Jakarta ValidationHandler channel (R98-blocked)

Split out from R12 (error-handling-parity)'s "Test fixture updates for source-direct dispatch" bullet. R12 lifted ValidationHandler channel support and the Jakarta pre-execution validation step into the per-fetcher catch path.

The original blocker, R94 (emit-input-records), has shipped: each SDL input type is now emitted as a graphitron-internal Java class under <outputPackage>.inputs, the fetcher boundary coerces the incoming map via fromMap, and the validator pre-step walks the typed instance instead of a raw Map. What remains missing is constraint content: the emitted classes carry no Jakarta annotations, so validator.validate(<typed>) still returns no violations and the wire path is a no-op end-to-end. Attaching constraints (SDL validation directives merged into a ConstraintSet, registered programmatically via ConstraintMapping) is R98 (multi-source-input-validation). Per R94’s Done note, this fixture picks up the live invalid-input round-trip the moment R98 ships its first SDL constraint; at that point it becomes a one-line addition to the sakila execute-tier driver.

Scope: an @service mutation whose input declares constraints (via whatever surface R98 lands), an @error type with {handler: VALIDATION}, and an execute-tier driver covering the violation path (constraint produces a typed error via ConstraintViolations.toGraphQLError) and the happy path. Pipeline-tier coverage of the emit shape is already pinned in TypeFetcherGeneratorTest and FetcherPipelineTest.