feat: support multiple households in calculation endpoints (#73)

* feat: support multiple households in calculation endpoints

Changed entity inputs from single dicts to lists of dicts, enabling
callers to submit multiple households with entity relational dataframes.
The caller specifies {entity}_id fields in each entity and
person_{entity}_id fields in people to link them together.

For simple single-household requests, IDs can be omitted and all people
will default to entity 0.

Co-Authored-By: Claude <noreply@anthropic.com>

* fix: route paths and test fixtures for aggregate endpoints

- Changed route paths from "/" to "" to avoid trailing slash mismatch
- Added simulation_id fixture to create test dependencies
- Updated tests to use the fixture instead of random UUIDs

Co-Authored-By: Claude <noreply@anthropic.com>

* fix: mock Modal functions in aggregate tests

---------

Co-authored-by: Claude <noreply@anthropic.com>

Author: nikhilwoodruff

src/policyengine_api/api/change_aggregates.py

@@ -83,7 +83,7 @@ def _trigger_change_aggregate_computation(
     )
 
 
-@router.post("/", response_model=List[ChangeAggregateRead])
+@router.post("", response_model=List[ChangeAggregateRead])
 def create_change_aggregates(
     outputs: List[ChangeAggregateCreate],
     background_tasks: BackgroundTasks,
@@ -128,7 +128,7 @@ def create_change_aggregates(
     return db_outputs
 
 
-@router.get("/", response_model=List[ChangeAggregateRead])
+@router.get("", response_model=List[ChangeAggregateRead])
 def list_change_aggregates(session: Session = Depends(get_session)):
     """List all change aggregates."""
     outputs = session.exec(select(ChangeAggregate)).all()