core/agent+cli: separate last-request usage from billed turn totals

Tool-loop rounds re-send history, so summed prompt tokens are billing
not context size. Track last_request_usage and label turn/session as
billed; clarify /status and end-of-turn lines.

tests/test_agent/test_loop.py

@@ -230,12 +230,51 @@ async def test_chat_agent_accumulates_session_usage() -> None:
     agent = ChatAgent(client, model="m", stream=False)
     _ = [e async for e in agent.run("one")]
     assert agent.last_turn_usage.total_tokens == 6
+    assert agent.last_request_usage.total_tokens == 6
+    assert agent.last_turn_rounds == 1
     assert agent.session_usage.total_tokens == 6
     _ = [e async for e in agent.run("two")]
     assert agent.last_turn_usage.total_tokens == 10
+    assert agent.last_request_usage.total_tokens == 10
     assert agent.session_usage.total_tokens == 16
 
 
+async def test_chat_agent_turn_usage_sums_tool_rounds() -> None:
+    """Multi-round tool loop: turn usage is billing sum; last_request is final call."""
+
+    @tool
+    def ping() -> str:
+        return "pong"
+
+    registry = ToolRegistry([ping])
+    client = ScriptedClient(
+        [
+            _response(
+                AssistantChatMessage(
+                    content="",
+                    tool_calls=[
+                        AssistantFunctionToolCall(
+                            id="1",
+                            function=AssistantFunctionTool(name="ping", arguments="{}"),
+                        )
+                    ],
+                ),
+                usage={"prompt_tokens": 100, "completion_tokens": 5, "total_tokens": 105},
+            ),
+            _response(
+                AssistantChatMessage(content="done"),
+                usage={"prompt_tokens": 200, "completion_tokens": 3, "total_tokens": 203},
+            ),
+        ]
+    )
+    agent = ChatAgent(client, model="m", tools=registry, stream=False)
+    _ = [e async for e in agent.run("go")]
+    assert agent.last_turn_rounds == 2
+    assert agent.last_request_usage.prompt_tokens == 200
+    assert agent.last_turn_usage.prompt_tokens == 300
+    assert agent.last_turn_usage.total_tokens == 308
+
+
 async def test_stream_yields_deltas_incrementally() -> None:
     """Text deltas are yielded as chunks arrive, not only after the full stream."""
 

tests/test_agent/test_usage.py

@@ -97,3 +97,5 @@ def test_format_line_marks_estimate() -> None:
     assert "(est)" in line
     mixed = TokenUsage(prompt_tokens=1, completion_tokens=0, total_tokens=1, source="mixed").format_line()
     assert "(api+est)" in mixed
+    billed = TokenUsage(prompt_tokens=1, completion_tokens=0, total_tokens=1).format_line(billed=True)
+    assert billed.startswith("billed tokens")