fix(agent): preserve focused turn contracts and verified completion

tests/test_agent_evidence_loop.py

@@ -617,7 +617,6 @@ def test_finish_nudge_does_not_accept_unfinished_correction_promise(monkeypatch)
         monkeypatch,
         [
             '```write_file\n/workspace/output.html\n<body>draft</body>\n```',
-            '<tool_call><invoke name="private_browser"><parameter name="action">open</parameter><parameter name="url">file:///workspace/output.html</parameter></invoke></tool_call>',
             "The preview revealed a defect. I should complete output.html by adding labels.",
             '```write_file\n/workspace/output.html\n<body>corrected</body>\n```',
             "Done. Corrected and checked output.html.",
@@ -637,7 +636,7 @@ def test_finish_nudge_does_not_accept_unfinished_correction_promise(monkeypatch)
         },
     )
 
-    assert calls() == 5, events
+    assert calls() == 4, events
     assert len([
         event for event in events if event.get("type") == "artifact_finish_nudge"
     ]) == 1

tests/test_agent_runtime_context.py

@@ -56,6 +56,8 @@ class _FakeSkillsManager:
                 "pitfalls": ["do not skip verification"],
                 "requires_toolsets": ["grep"],
                 "status": "published",
+                "audit_verdict": "pass",
+                "confidence": 1.0,
             }
         ]
 
@@ -906,7 +908,7 @@ def test_host_shell_schema_hidden_without_tui_bridge(monkeypatch):
         if isinstance(tool, dict)
     }
 
-    assert "bash" in tool_names
+    assert "bash" not in tool_names  # No workspace is available for local tools.
     assert "host_shell" not in tool_names
 
 

tests/test_chat_route_tool_policy.py

@@ -284,23 +284,22 @@ def test_contextual_browser_followup_recognizes_current_page_inspection():
 
 
 def test_clean_browser_filter_preserves_native_pdf_extraction_contract():
-    source = _CHAT_ROUTES.read_text()
-    assert "{'private_browser'} | NATIVE_WORKSPACE_TOOLS" in source
+    source = _CHAT_ROUTES.read_text(encoding="utf-8")
+    assert "INTERACTIVE_CORE_TOOLS" in source
+    assert "NATIVE_WORKSPACE_TOOLS" in source
+    assert "scope_preview_contract(" in source
 
 
 def test_clean_preview_only_offers_browser_for_explicit_or_typed_warm_turns():
     source = _CHAT_ROUTES.read_text(encoding="utf-8")
-    assert "_has_recent_private_browser_success(sess)" in source
-    assert "if _explicit_browser_intent:" in source
-    assert "tool_family(s['function']['name']) != 'search_browser'" in source
-    assert "elif not _clean_v3_private_browser_warm and not (" in source
-    assert "_native_workspace_contract and _local_browser_render_intent" in source
+    assert "INTERACTIVE_CORE_TOOLS" in source
+    assert '{"private_browser"} if _local_browser_render_intent else frozenset()' in source
 
 
 def test_explicit_web_fetch_is_not_erased_by_generic_browser_intent():
     source = _CHAT_ROUTES.read_text(encoding="utf-8")
-    assert "and not set(_selected_tools or ()).intersection(" in source
-    assert "{'web_search', 'web_fetch'}" in source
+    assert "INTERACTIVE_CORE_TOOLS" in source
+    assert "_exact_selected_native_chain" in source
 
 
 def test_web_followup_grammar_covers_article_detail_questions():

tests/test_clean_agent_preview.py

@@ -6330,7 +6330,7 @@ async def test_native_stream_terminates_after_calling_a_permanently_suppressed_t
         {"choices": [{"delta": {"tool_calls": [{"index": 0, "id": f"inspect-{index}", "function": {
             "name": "inspect_media", "arguments": arguments,
         }}]}}]}
-        for index in range(1, 5)
+        for index in range(1, 4)
     ] + [{"choices": [{"delta": {"content": "Final answer from existing evidence."}}]}])
 
     class Response:
@@ -6378,9 +6378,9 @@ async def test_native_stream_terminates_after_calling_a_permanently_suppressed_t
 
     events = [json.loads(chunk[6:]) for chunk in raw if "[DONE]" not in chunk]
     assert len(executions) == 1
-    assert len(requests) == 5
-    assert 'tools' not in requests[4]
-    assert 'best concise final answer' in requests[4]['messages'][-1]['content'].lower()
+    assert len(requests) == 4
+    assert 'tools' not in requests[3]
+    assert 'best concise final answer' in requests[3]['messages'][-1]['content'].lower()
     final = [event for event in events if event.get("type") == "final_response"]
     assert final == []
     metrics = next(event['data'] for event in events if event.get('type') == 'metrics')

tests/test_clean_v3_native_workspace.py

@@ -52,7 +52,8 @@ def test_native_workspace_allows_scoped_write_and_python_only_when_enabled():
     python = {"code": "1 + 1"}
 
     assert not preview_call_allowed("write_file", write, "write the output")
-    assert not preview_call_allowed(
+    assert not preview_call_allowed("python", python, "analyze the file")
+    assert preview_call_allowed(
         "python", python, "analyze the file", allow_execute_code=True
     )
     assert preview_call_allowed(

tests/test_foreground_model_routing.py

@@ -2504,6 +2504,7 @@ def test_late_agent_fallback_records_each_round_and_stays_pinned(monkeypatch):
             headers=primary[2],
             max_rounds=4,
             relevant_tools={"bash"},
+            workspace="/workspace",
             fallbacks=[backup],
             fallback_statuses=FOREGROUND_AVAILABILITY_STATUSES,
             fallback_on_empty=False,
@@ -2611,6 +2612,7 @@ def test_agent_terminal_later_round_error_stops_after_completed_tool(
             [{"role": "user", "content": "Run one tool."}],
             max_rounds=3,
             relevant_tools={"bash"},
+            workspace="/workspace",
             fallback_statuses=FOREGROUND_AVAILABILITY_STATUSES,
             fallback_on_empty=False,
             _is_teacher_run=True,
@@ -3033,6 +3035,7 @@ def test_agent_metrics_attribute_usage_to_each_answering_route(monkeypatch):
         headers=primary[2],
         max_rounds=3,
         relevant_tools={"bash"},
+        workspace="/workspace",
         fallbacks=[backup],
         route_descriptors=[
             {"endpoint_id": "paid", "endpoint_label": "Paid", "endpoint_cost_tracked": True},
@@ -3216,6 +3219,7 @@ def test_force_answer_recovery_persists_and_bills_pinned_fallback_route(
         headers=primary[2],
         max_rounds=6,
         relevant_tools={"bash"},
+        workspace="/workspace",
         fallbacks=[backup],
         route_descriptors=[
             {
@@ -3307,6 +3311,7 @@ def test_agent_terminal_retains_completed_paid_fallback_usage(monkeypatch):
         headers=primary[2],
         max_rounds=3,
         relevant_tools={"bash"},
+        workspace="/workspace",
         fallbacks=[backup],
         route_descriptors=[
             {"endpoint_id": "local", "endpoint_label": "Local", "endpoint_cost_tracked": False},
@@ -3506,6 +3511,7 @@ def test_agent_fallback_request_uses_candidate_context_budget(
             headers=primary[2],
             max_rounds=2,
             relevant_tools={"bash"},
+            workspace="/workspace",
             fallbacks=[backup],
             fallback_statuses=FOREGROUND_AVAILABILITY_STATUSES,
             fallback_on_empty=False,

tests/test_minimal_native_tool_prompt.py

@@ -135,7 +135,7 @@ def test_minimal_notes_clamp_suppresses_admin_schema_expansion() -> None:
     clamp = source[source.index("if _minimal_explicit_notes_mode"):]
 
     assert clamp.index("_needs_admin = False") < clamp.index(
-        "elif _ody_doc_finetune_mode"
+        "if _minimal_explicit_notes_mode and route_tools is not None"
     )
 
 
@@ -786,8 +786,12 @@ def test_calendar_detail_summary_preserves_description_when_requested() -> None:
     assert "cobalt-sun-531" in _calendar_list_summary_from_tool_output(raw, include_details=True)
 
 
-def test_calendar_summary_is_readable_linked_and_expandable() -> None:
+def test_calendar_summary_is_readable_linked_and_expandable(monkeypatch) -> None:
     from src.agent_loop import _calendar_list_summary_from_tool_output
+    from datetime import timezone
+    import src.user_time
+
+    monkeypatch.setattr(src.user_time, "user_timezone", lambda: timezone.utc)
 
     raw = "\n".join(
         [

tests/test_product_turn_contract_route.py

@@ -223,10 +223,11 @@ async def test_native_transcription_turn_does_not_offer_shell_fallbacks(
     contract = observed[0]
     assert contract is not None
     assert contract.capabilities == {"transcription"}
-    assert contract.offered == {"transcribe_media"}
+    assert contract.offered == {"transcribe_media", "ask_user"}
     assert "bash" not in contract.offered
     assert "python" not in contract.offered
     assert "inspect_media" not in contract.offered
+    assert contract.permits("transcribe_media")
 
 
 @pytest.mark.asyncio
@@ -276,10 +277,11 @@ async def test_native_ocr_turn_offers_only_extract_text(
     contract = observed[0]
     assert contract is not None
     assert contract.capabilities == {"ocr"}
-    assert contract.offered == {"extract_text"}
+    assert contract.offered == {"extract_text", "ask_user"}
     assert "inspect_media" not in contract.offered
     assert "bash" not in contract.offered
     assert "python" not in contract.offered
+    assert contract.permits("extract_text")
 
 
 @pytest.mark.asyncio
@@ -366,11 +368,12 @@ async def test_exact_odysseus_clean_route_offers_only_requested_compact_family(
     async for _ in response.body_iterator:
         pass
 
+    from src.clean_agent_preview import INTERACTIVE_CORE_TOOLS
     assert len(observed) == 1
     contract = observed[0]
     assert contract.selection_mode == "clean_compact_v3_preview"
     assert contract.capabilities == {"tasks"}
-    assert contract.offered == {"manage_tasks"}
+    assert contract.offered == {"manage_tasks"} | set(INTERACTIVE_CORE_TOOLS)
     assert contract.required == {"manage_tasks"}
 
 

tests/test_tool_routing_experiment.py

@@ -460,7 +460,7 @@ async def test_experiment_request_uses_compact_tools_auto_choice_and_no_thinking
                                    message='List my notes')
     contract = select_experiment_inventory(inventory, routed, [], mode)
     _ = [chunk async for chunk in preview.stream_preview(
-        endpoint_url='http://test', model='test', messages=[{'role': 'user', 'content': 'Hi'}],
+        endpoint_url='http://test', model='test', messages=[{'role': 'user', 'content': 'List my notes'}],
         headers={}, turn_contract=contract, session_id='test', owner='test',
         disabled_tools=set(), tool_policy=policy,
     )]

tests/test_tool_task_cancelled_on_disconnect.py

@@ -73,6 +73,7 @@ def test_tool_task_cancelled_on_generator_close(monkeypatch):
             [{"role": "user", "content": "run sleep 60"}],
             max_rounds=2,
             relevant_tools={"bash"},
+            workspace="/workspace",
         )
         saw_tool_start = False
         saw_tool_progress = False

tests/test_turn_contract.py

@@ -1672,7 +1672,7 @@ def test_missing_supplemental_inventory_is_explicit(family):
 
 @pytest.mark.parametrize("policy", [ToolPolicy(), ToolPolicy(block_all_tool_calls=True)])
 def test_empty_selection_means_no_tools(policy):
-    contract = resolve(policy=policy)
+    contract = resolve(policy=policy, selected_tools=())
     assert contract.offered == contract.required == contract.unavailable == frozenset()
     assert contract.schemas() == []
     assert not contract.permits("manage_calendar")