fix(agent): refresh llm runtime for background tasks

nanobot/agent/loop.py

@@ -41,6 +41,7 @@ from nanobot.utils.document import extract_documents
 from nanobot.utils.helpers import image_placeholder_text
 from nanobot.utils.helpers import truncate_text as truncate_text_fn
 from nanobot.utils.image_generation_intent import image_generation_prompt
+from nanobot.utils.llm_runtime import LLMRuntime
 from nanobot.utils.runtime import EMPTY_FINAL_RESPONSE_MESSAGE
 from nanobot.utils.session_attachments import merge_turn_media_into_last_assistant
 from nanobot.utils.webui_turn_helpers import (
@@ -138,6 +139,11 @@ class AgentLoop:
     def tool_names(self) -> list[str]:
         return self.tools.tool_names
 
+    def llm_runtime(self) -> LLMRuntime:
+        """Return the current provider/model pair owned by this loop."""
+        self._refresh_provider_snapshot()
+        return LLMRuntime(self.provider, self.model)
+
     _RUNTIME_CHECKPOINT_KEY = "runtime_checkpoint"
     _PENDING_USER_TURN_KEY = "pending_user_turn"
 
@@ -1296,8 +1302,7 @@ class AgentLoop:
         self._webui_turns.capture_title_context(
             ctx.session_key,
             ctx.msg,
-            self.provider,
-            self.model,
+            self.llm_runtime(),
         )
 
         ctx.initial_messages = self._build_initial_messages(

nanobot/cli/commands.py

@@ -914,8 +914,7 @@ def _run_gateway(
     hb_cfg = config.gateway.heartbeat
     heartbeat = HeartbeatService(
         workspace=config.workspace_path,
-        provider=agent.provider,
-        model=agent.model,
+        llm_runtime=agent.llm_runtime,
         on_execute=on_heartbeat_execute,
         on_notify=on_heartbeat_notify,
         interval_s=hb_cfg.interval_s,

nanobot/heartbeat/service.py

@@ -4,12 +4,12 @@ from __future__ import annotations
 
 import asyncio
 from pathlib import Path
-from typing import TYPE_CHECKING, Any, Callable, Coroutine
+from typing import Any, Callable, Coroutine
 
 from loguru import logger
 
-if TYPE_CHECKING:
-    from nanobot.providers.base import LLMProvider
+from nanobot.providers.base import LLMProvider
+from nanobot.utils.llm_runtime import LLMRuntime, LLMRuntimeResolver, static_llm_runtime
 
 _HEARTBEAT_TOOL = [
     {
@@ -53,17 +53,21 @@ class HeartbeatService:
     def __init__(
         self,
         workspace: Path,
-        provider: LLMProvider,
-        model: str,
+        provider: LLMProvider | None = None,
+        model: str | None = None,
         on_execute: Callable[[str], Coroutine[Any, Any, str]] | None = None,
         on_notify: Callable[[str], Coroutine[Any, Any, None]] | None = None,
         interval_s: int = 30 * 60,
         enabled: bool = True,
         timezone: str | None = None,
+        llm_runtime: LLMRuntimeResolver | None = None,
     ):
         self.workspace = workspace
-        self.provider = provider
-        self.model = model
+        if llm_runtime is None:
+            if provider is None or model is None:
+                raise ValueError("HeartbeatService requires either llm_runtime or provider/model")
+            llm_runtime = static_llm_runtime(provider, model)
+        self._llm_runtime = llm_runtime
         self.on_execute = on_execute
         self.on_notify = on_notify
         self.interval_s = interval_s
@@ -91,7 +95,9 @@ class HeartbeatService:
         """
         from nanobot.utils.helpers import current_time_str
 
-        response = await self.provider.chat_with_retry(
+        llm = self._llm_runtime()
+
+        response = await llm.provider.chat_with_retry(
             messages=[
                 {"role": "system", "content": "You are a heartbeat agent. Call the heartbeat tool to report your decision."},
                 {"role": "user", "content": (
@@ -101,7 +107,7 @@ class HeartbeatService:
                 )},
             ],
             tools=_HEARTBEAT_TOOL,
-            model=self.model,
+            model=llm.model,
         )
 
         if not response.should_execute_tools:
@@ -214,8 +220,9 @@ class HeartbeatService:
                     )
                     return
 
+                llm = self._llm_runtime()
                 should_notify = await evaluate_response(
-                    response, tasks, self.provider, self.model,
+                    response, tasks, llm.provider, llm.model,
                 )
                 if should_notify and self.on_notify:
                     logger.info("Heartbeat: completed, delivering response")

nanobot/utils/llm_runtime.py

@@ -0,0 +1,22 @@
+"""Small helpers for passing the active LLM provider/model together."""
+
+from __future__ import annotations
+
+from collections.abc import Callable
+from dataclasses import dataclass
+
+from nanobot.providers.base import LLMProvider
+
+
+@dataclass(frozen=True)
+class LLMRuntime:
+    provider: LLMProvider
+    model: str
+
+
+LLMRuntimeResolver = Callable[[], LLMRuntime]
+
+
+def static_llm_runtime(provider: LLMProvider, model: str) -> LLMRuntimeResolver:
+    runtime = LLMRuntime(provider=provider, model=model)
+    return lambda: runtime

nanobot/utils/webui_turn_helpers.py

@@ -20,6 +20,7 @@ from nanobot.providers.base import LLMProvider
 from nanobot.session.goal_state import goal_state_ws_blob
 from nanobot.session.manager import Session, SessionManager
 from nanobot.utils.helpers import truncate_text
+from nanobot.utils.llm_runtime import LLMRuntime
 
 WEBUI_SESSION_METADATA_KEY = "webui"
 WEBUI_TITLE_METADATA_KEY = "title"
@@ -31,7 +32,6 @@ TITLE_GENERATION_REASONING_EFFORT = "none"
 # Wall-clock turn start per ``chat_id`` (websocket only). Survives browser refresh while the
 # gateway process stays up; cleared on idle/stop and implicitly dropped on restart.
 _WEBSOCKET_TURN_WALL_STARTED_AT: dict[str, float] = {}
-TitleContext = tuple[LLMProvider, str]
 
 
 def mark_webui_session(session: Session, metadata: dict[str, Any]) -> bool:
@@ -241,17 +241,16 @@ class WebuiTurnCoordinator:
     bus: MessageBus
     sessions: SessionManager
     schedule_background: Callable[[Awaitable[None]], None]
-    _title_contexts: dict[str, TitleContext] = field(default_factory=dict)
+    _title_contexts: dict[str, LLMRuntime] = field(default_factory=dict)
 
     def capture_title_context(
         self,
         session_key: str,
         msg: InboundMessage,
-        provider: LLMProvider,
-        model: str,
+        llm: LLMRuntime,
     ) -> None:
         if msg.channel == "websocket" and msg.metadata.get("webui") is True:
-            self._title_contexts[session_key] = (provider, model)
+            self._title_contexts[session_key] = llm
 
     def discard(self, session_key: str) -> None:
         self._title_contexts.pop(session_key, None)
@@ -287,19 +286,16 @@ class WebuiTurnCoordinator:
         if msg.metadata.get("webui") is not True or title_context is None:
             return
 
-        title_provider, title_model = title_context
-
         async def _generate_title_and_notify(
-            provider: LLMProvider = title_provider,
-            model: str = title_model,
+            title_llm: LLMRuntime = title_context,
         ) -> None:
             generated = await maybe_generate_webui_title_after_turn(
                 channel=msg.channel,
                 metadata=msg.metadata,
                 sessions=self.sessions,
                 session_key=session_key,
-                provider=provider,
-                model=model,
+                provider=title_llm.provider,
+                model=title_llm.model,
             )
             if generated:
                 await self.bus.publish_outbound(OutboundMessage(

tests/agent/test_heartbeat_service.py

@@ -4,6 +4,7 @@ import pytest
 
 from nanobot.heartbeat.service import HeartbeatService
 from nanobot.providers.base import LLMProvider, LLMResponse, ToolCallRequest
+from nanobot.utils.llm_runtime import LLMRuntime
 
 
 class DummyProvider(LLMProvider):
@@ -11,9 +12,11 @@ class DummyProvider(LLMProvider):
         super().__init__()
         self._responses = list(responses)
         self.calls = 0
+        self.models: list[str | None] = []
 
     async def chat(self, *args, **kwargs) -> LLMResponse:
         self.calls += 1
+        self.models.append(kwargs.get("model"))
         if self._responses:
             return self._responses.pop(0)
         return LLMResponse(content="", tool_calls=[])
@@ -215,6 +218,51 @@ async def test_tick_suppresses_when_evaluator_says_no(tmp_path, monkeypatch) ->
     assert notified == []
 
 
+def test_tick_uses_runtime_provider_and_model(tmp_path, monkeypatch) -> None:
+    """Preset changes must apply to heartbeat decision and post-run evaluation."""
+    (tmp_path / "HEARTBEAT.md").write_text("- [ ] check runtime model", encoding="utf-8")
+
+    runtime_provider = DummyProvider([
+        LLMResponse(
+            content="",
+            tool_calls=[
+                ToolCallRequest(
+                    id="hb_1",
+                    name="heartbeat",
+                    arguments={"action": "run", "tasks": "check runtime model"},
+                )
+            ],
+        ),
+    ])
+    runtime_model = "openai/gpt-4.1"
+
+    executed: list[str] = []
+    evaluated: list[tuple[LLMProvider, str]] = []
+
+    async def _on_execute(tasks: str) -> str:
+        executed.append(tasks)
+        return "runtime model produced a user-facing update"
+
+    async def _eval_capture(response, tasks, provider, model):
+        evaluated.append((provider, model))
+        return False
+
+    service = HeartbeatService(
+        workspace=tmp_path,
+        llm_runtime=lambda: LLMRuntime(runtime_provider, runtime_model),
+        on_execute=_on_execute,
+    )
+
+    monkeypatch.setattr("nanobot.utils.evaluator.evaluate_response", _eval_capture)
+
+    asyncio.run(service._tick())
+
+    assert runtime_provider.calls == 1
+    assert runtime_provider.models == [runtime_model]
+    assert executed == ["check runtime model"]
+    assert evaluated == [(runtime_provider, runtime_model)]
+
+
 @pytest.mark.asyncio
 async def test_decide_retries_transient_error_then_succeeds(tmp_path, monkeypatch) -> None:
     provider = DummyProvider([
@@ -286,4 +334,3 @@ async def test_decide_prompt_includes_current_time(tmp_path) -> None:
     user_msg = captured_messages[1]
     assert user_msg["role"] == "user"
     assert "Current Time:" in user_msg["content"]
-

tests/agent/test_loop_save_turn.py

@@ -10,14 +10,16 @@ from nanobot.bus.events import InboundMessage
 from nanobot.bus.queue import MessageBus
 from nanobot.providers.base import LLMResponse
 from nanobot.session.goal_state import GOAL_STATE_KEY
-from nanobot.session.manager import Session
+from nanobot.session.manager import Session, SessionManager
 from nanobot.utils.webui_turn_helpers import (
     TITLE_GENERATION_MAX_TOKENS,
     TITLE_GENERATION_REASONING_EFFORT,
     WEBUI_SESSION_METADATA_KEY,
     WEBUI_TITLE_METADATA_KEY,
+    WebuiTurnCoordinator,
     maybe_generate_webui_title,
 )
+from nanobot.utils.llm_runtime import LLMRuntime
 
 
 def _mk_loop() -> AgentLoop:
@@ -35,6 +37,22 @@ def _make_full_loop(tmp_path: Path) -> AgentLoop:
     return AgentLoop(bus=MessageBus(), provider=provider, workspace=tmp_path, model="test-model")
 
 
+def test_agent_loop_llm_runtime_reflects_current_provider_and_model(tmp_path: Path) -> None:
+    loop = _make_full_loop(tmp_path)
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is loop.provider
+    assert runtime.model == "test-model"
+
+    next_provider = MagicMock()
+    loop.provider = next_provider
+    loop.model = "next-model"
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is next_provider
+    assert runtime.model == "next-model"
+
+
 @pytest.mark.asyncio
 async def test_generate_webui_title_only_for_marked_webui_sessions(tmp_path: Path) -> None:
     loop = _make_full_loop(tmp_path)
@@ -111,6 +129,55 @@ async def test_generate_webui_title_ignores_command_only_sessions(tmp_path: Path
     loop.provider.chat_with_retry.assert_not_awaited()
 
 
+def test_webui_title_update_uses_captured_llm_runtime(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    bus = MessageBus()
+    sessions = SessionManager(tmp_path)
+    scheduled: list[object] = []
+    captured: dict[str, object] = {}
+
+    async def fake_title_after_turn(**kwargs: object) -> bool:
+        captured.update(kwargs)
+        return False
+
+    monkeypatch.setattr(
+        "nanobot.utils.webui_turn_helpers.maybe_generate_webui_title_after_turn",
+        fake_title_after_turn,
+    )
+    coordinator = WebuiTurnCoordinator(
+        bus=bus,
+        sessions=sessions,
+        schedule_background=lambda coro: scheduled.append(coro),
+    )
+    provider = MagicMock()
+    msg = InboundMessage(
+        channel="websocket",
+        sender_id="u1",
+        chat_id="chat1",
+        content="say hello",
+        metadata={"webui": True},
+    )
+
+    coordinator.capture_title_context(
+        "websocket:chat1",
+        msg,
+        LLMRuntime(provider, "turn-model"),
+    )
+    asyncio.run(coordinator.handle_turn_end(
+        msg,
+        session_key="websocket:chat1",
+        latency_ms=None,
+    ))
+
+    assert len(scheduled) == 1
+    asyncio.run(scheduled[0])  # type: ignore[arg-type]
+
+    assert captured["provider"] is provider
+    assert captured["model"] == "turn-model"
+
+
 def test_save_turn_skips_multimodal_user_when_only_runtime_context() -> None:
     loop = _mk_loop()
     session = Session(key="test:runtime-only")

tests/agent/test_runtime_refresh.py

@@ -47,3 +47,28 @@ def test_provider_refresh_updates_all_model_dependents(tmp_path: Path) -> None:
     assert loop.dream.provider is new_provider
     assert loop.dream.model == "new-model"
     assert loop.dream._runner.provider is new_provider
+
+
+def test_llm_runtime_refreshes_provider_snapshot(tmp_path: Path) -> None:
+    old_provider = _provider("old-model")
+    new_provider = _provider("new-model", max_tokens=456)
+    loop = AgentLoop(
+        bus=MessageBus(),
+        provider=old_provider,
+        workspace=tmp_path,
+        model="old-model",
+        context_window_tokens=1000,
+        provider_snapshot_loader=lambda: ProviderSnapshot(
+            provider=new_provider,
+            model="new-model",
+            context_window_tokens=2000,
+            signature=("new-model",),
+        ),
+    )
+
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is new_provider
+    assert runtime.model == "new-model"
+    assert loop.provider is new_provider
+    assert loop.runner.provider is new_provider

tests/cli/test_commands.py

@@ -1170,6 +1170,7 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
             self.model = "test-model"
             self.provider = kwargs.get("provider", object())
             self.tools = {}
+            seen["agent"] = self
 
         async def process_direct(self, *_args, **_kwargs):
             return OutboundMessage(
@@ -1218,6 +1219,11 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
     assert isinstance(cron, _FakeCron)
     assert cron.on_job is not None
 
+    runtime_provider = object()
+    agent = seen["agent"]
+    agent.provider = runtime_provider
+    agent.model = "runtime-model"
+
     job = CronJob(
         id="cron-1",
         name="stretch",
@@ -1233,8 +1239,8 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
 
     assert response == "Time to stretch."
     assert seen["response"] == "Time to stretch."
-    assert seen["provider"] is provider
-    assert seen["model"] == "test-model"
+    assert seen["provider"] is runtime_provider
+    assert seen["model"] == "runtime-model"
     assert seen["task_context"] == (
         "The scheduled time has arrived. Deliver this reminder to the user now, "
         "as a brief and natural message in their language. Speak directly to them — "