Blueprint

Compose two blueprints into a single composed blueprint.

This is the core operation for Phase 3: combining two independent module blueprints into a single ComposedBlueprint with:

• Unioned owned schemas (O1 ++ O2)
• Unioned transaction sets (T1 ++ T2)
• Merged reducer logic with module-based routing

The composed blueprint requires evidence that the combined schema has unique table names (UniqueNames[O1 ++ O2]).

Reducer routing strategy:

• Transactions MUST implement ModuleRoutedTx (enforced by RoutedStateReducer0 type)
• The moduleId.path is module-relative (MName *: SubPath)
• The reducer routes based on the first segment matching M1 or M2
• Type safety is compile-time: non-routed transactions will not compile

IMPORTANT: ModuleId is always module-relative. When a blueprint is mounted at a path, the mount path is NOT prepended to transaction moduleIds. Full paths (mountPath ++ moduleId.path) are only constructed at system edges for telemetry or logging.

Phase 5.6 UPGRADE: Now supports modules with non-empty Needs. The providers are merged using TablesProvider.merge, which requires that the dependency schemas are disjoint (DisjointSchemas[N1, N2]). This prevents ambiguous table lookups while allowing flexible composition of dependent modules.

Mount a blueprint at a path composed from base and sub paths.

This is a convenience helper for Phase 3 that simplifies mounting a blueprint at a nested path.

Type inference: Only Base and Sub need to be specified explicitly. All other types (F, MName, Owns, Needs, Txs) are inferred from the blueprint parameter.