feat(models): add StepResult and ResetResponse models

models.py

@@ -47,3 +47,11 @@ class EpisodeState(BaseModel):
     max_turns:     int
     attacks_so_far:int
     is_active:     bool
+
+class StepResult(BaseModel):
+    observation: RedTeamObservation
+    reward:      float
+
+class ResetResponse(BaseModel):
+    observation: RedTeamObservation
+    episode_id:  str