Harden SBOM retries and align hub evidence

Assistant: codex
Assistant-Model: gpt-5.6-sol
Assistant-Session: 01a028de-e2c8-7732-8521-46a7fc5db82f

src/activity_core/worker.py

@@ -33,6 +33,7 @@ from temporalio.runtime import PrometheusConfig, Runtime, TelemetryConfig
 from temporalio.worker import Worker
 
 from activity_core.activities import (
+    apply_sbom_catchup,
     emit_tasks,
     evaluate_instructions,
     evaluate_rules,
@@ -104,6 +105,7 @@ async def run() -> None:
         activities=[
             load_activity_definition,
             resolve_context,
+            apply_sbom_catchup,
             log_run,
             evaluate_rules,
             evaluate_instructions,