refactor: update all imports for new directory structure

Update import paths in server/, tests/, and root __init__.py to
reference modules in their new locations (core/, evaluation/,
runners/, scenarios/).

__init__.py

@@ -1,7 +1,7 @@
 """ChargebackOps OpenEnv package."""
 
-from .client import ChargebackOpsEnv
-from .models import (
+from .core.client import ChargebackOpsEnv
+from .core.models import (
     BaselineRunResult,
     ChargebackOpsAction,
     ChargebackOpsObservation,

server/app.py

@@ -13,30 +13,30 @@ except Exception as exc:  # pragma: no cover
     ) from exc
 
 try:
-    from ..baseline_runner import run_baseline
-    from ..episode_store import get_report
-    from ..inference import run_inference
-    from ..models import (
+    from ..runners.baseline_runner import run_baseline
+    from ..core.episode_store import get_report
+    from ..runners.inference import run_inference
+    from ..core.models import (
         BaselineRunResult,
         ChargebackOpsAction,
         ChargebackOpsObservation,
         TasksResponse,
         TaskSummary,
     )
-    from ..simulation import list_tasks
+    from ..scenarios.simulation import list_tasks
     from .chargeback_ops_environment import ChargebackOpsEnvironment
 except ImportError:  # pragma: no cover
-    from baseline_runner import run_baseline
-    from episode_store import get_report
-    from inference import run_inference
-    from models import (
+    from runners.baseline_runner import run_baseline
+    from core.episode_store import get_report
+    from runners.inference import run_inference
+    from core.models import (
         BaselineRunResult,
         ChargebackOpsAction,
         ChargebackOpsObservation,
         TasksResponse,
         TaskSummary,
     )
-    from simulation import list_tasks
+    from scenarios.simulation import list_tasks
     from server.chargeback_ops_environment import ChargebackOpsEnvironment
 
 
@@ -95,9 +95,9 @@ def generate_tasks(
     """Generate parametric tasks from a seed for infinite scenario variety."""
 
     try:
-        from case_generator import generate_task_suite
+        from scenarios.case_generator import generate_task_suite
     except ImportError:  # pragma: no cover
-        from ..case_generator import generate_task_suite
+        from ..scenarios.case_generator import generate_task_suite
 
     suite = generate_task_suite(
         base_seed=seed, easy_count=easy, medium_count=medium, hard_count=hard,

server/chargeback_ops_environment.py

@@ -7,9 +7,9 @@ from uuid import uuid4
 from openenv.core.env_server.interfaces import Environment
 
 try:
-    from ..episode_store import record_report
-    from ..grading import grade_episode
-    from ..models import (
+    from ..core.episode_store import record_report
+    from ..evaluation.grading import grade_episode
+    from ..core.models import (
         ActionTraceItem,
         CaseQueueItem,
         CaseResolutionState,
@@ -20,11 +20,11 @@ try:
         PolicyView,
         VisibleCase,
     )
-    from ..simulation import ActionRecord, CaseProgress, InternalCase, get_task
+    from ..scenarios.simulation import ActionRecord, CaseProgress, InternalCase, get_task
 except ImportError:  # pragma: no cover
-    from episode_store import record_report
-    from grading import grade_episode
-    from models import (
+    from core.episode_store import record_report
+    from evaluation.grading import grade_episode
+    from core.models import (
         ActionTraceItem,
         CaseQueueItem,
         CaseResolutionState,
@@ -35,7 +35,7 @@ except ImportError:  # pragma: no cover
         PolicyView,
         VisibleCase,
     )
-    from simulation import ActionRecord, CaseProgress, InternalCase, get_task
+    from scenarios.simulation import ActionRecord, CaseProgress, InternalCase, get_task
 
 
 class ChargebackOpsEnvironment(

tests/test_agent_audit.py

@@ -1,4 +1,4 @@
-from agent_brutal_audit import aggregate_results, run_episode
+from evaluation.agent_brutal_audit import aggregate_results, run_episode
 
 
 def test_heuristic_beats_bad_on_generated_suite():

tests/test_api.py

@@ -1,5 +1,5 @@
-from inference import run_inference
-from models import ChargebackOpsAction
+from runners.inference import run_inference
+from core.models import ChargebackOpsAction
 from server.app import baseline, grader, root, tasks
 from server.chargeback_ops_environment import ChargebackOpsEnvironment
 

tests/test_env.py

@@ -1,5 +1,5 @@
-from case_generator import generate_task
-from models import ChargebackOpsAction
+from scenarios.case_generator import generate_task
+from core.models import ChargebackOpsAction
 from server.chargeback_ops_environment import ChargebackOpsEnvironment
 
 

tests/test_grader.py

@@ -1,6 +1,6 @@
-from grading import grade_episode
+from evaluation.grading import grade_episode
 from server.chargeback_ops_environment import ChargebackOpsEnvironment
-from simulation import get_task
+from scenarios.simulation import get_task
 
 
 def test_grade_episode_bounds():

tests/test_requirements.py

@@ -2,11 +2,11 @@ from __future__ import annotations
 
 from pathlib import Path
 
-from baseline_runner import _heuristic_pick, _obvious_next_action, candidate_actions
-from grading import grade_episode
-from models import ChargebackOpsAction
+from runners.baseline_runner import _heuristic_pick, _obvious_next_action, candidate_actions
+from evaluation.grading import grade_episode
+from core.models import ChargebackOpsAction
 from server.chargeback_ops_environment import ChargebackOpsEnvironment
-from simulation import get_task, list_tasks
+from scenarios.simulation import get_task, list_tasks
 
 
 def _run_heuristic_episode(task_id: str) -> tuple[float, float]:
@@ -162,7 +162,7 @@ def test_problem_statement_live_agent_budget_targets_real_branches():
 
 
 def test_problem_statement_inference_contract_exists():
-    content = Path("inference.py").read_text()
+    content = Path("runners/inference.py").read_text()
     assert "from openai import OpenAI" in content
     assert "API_BASE_URL" in content
     assert "MODEL_NAME" in content