refactor(agent): centralize runtime default mutations

nanobot/agent/model_runtime.py

@@ -3,6 +3,7 @@
 from __future__ import annotations
 
 from collections.abc import Callable, Mapping
+from dataclasses import replace
 
 from nanobot.agent import model_presets as preset_helpers
 from nanobot.config.schema import Config, ModelPresetConfig
@@ -31,6 +32,7 @@ class ModelRuntimeResolver:
         self._provider_snapshot_loader = provider_snapshot_loader
         self._preset_snapshot_loader = preset_snapshot_loader
         self._active_preset = initial_runtime.model_preset
+        self._tracks_provider_generation = initial_runtime.model_preset is None
         self._default_selection_signature = preset_helpers.default_selection_signature(
             initial_runtime.snapshot_signature
         )
@@ -56,6 +58,7 @@ class ModelRuntimeResolver:
         """Return the selected runtime, optionally refreshing the default source."""
         if refresh:
             self.refresh()
+            self._refresh_provider_generation()
         return self._runtime
 
     def resolve_snapshot(
@@ -77,6 +80,7 @@ class ModelRuntimeResolver:
         runtime = self.resolve_snapshot(snapshot, model_preset=model_preset)
         self._runtime = runtime
         self._active_preset = model_preset
+        self._tracks_provider_generation = model_preset is None
         self._default_selection_signature = preset_helpers.default_selection_signature(
             runtime.snapshot_signature
         )
@@ -98,8 +102,51 @@ class ModelRuntimeResolver:
         runtime = self.resolve_preset(name)
         self._runtime = runtime
         self._active_preset = runtime.model_preset
+        self._tracks_provider_generation = False
         return runtime
 
+    def select_model(self, model: str) -> LLMRuntime:
+        """Change the default model without reconstructing downstream consumers."""
+        if not isinstance(model, str) or not model.strip():
+            raise ValueError("model must be a non-empty string")
+        self._runtime = replace(
+            self._runtime,
+            model=model.strip(),
+            model_preset=None,
+        )
+        self._active_preset = None
+        return self._runtime
+
+    def select_context_window(self, context_window_tokens: int) -> LLMRuntime:
+        """Change the default context limit for future admissions."""
+        if not isinstance(context_window_tokens, int) or isinstance(
+            context_window_tokens,
+            bool,
+        ):
+            raise TypeError("context_window_tokens must be an integer")
+        self._runtime = replace(
+            self._runtime,
+            context_window_tokens=context_window_tokens,
+        )
+        return self._runtime
+
+    def _refresh_provider_generation(self) -> LLMRuntime | None:
+        """Adopt direct provider-default changes only for provider-backed defaults."""
+        if not self._tracks_provider_generation:
+            return None
+        runtime = self._runtime
+        captured = LLMRuntime.capture(
+            runtime.provider,
+            runtime.model,
+            context_window_tokens=runtime.context_window_tokens,
+            model_preset=runtime.model_preset,
+            snapshot_signature=runtime.snapshot_signature,
+        )
+        if captured.generation == runtime.generation:
+            return None
+        self._runtime = replace(runtime, generation=captured.generation)
+        return self._runtime
+
     def refresh(self) -> LLMRuntime | None:
         """Refresh configured defaults and return the replacement when changed."""
         if self._provider_snapshot_loader is None:
@@ -121,6 +168,7 @@ class ModelRuntimeResolver:
             return None
         self._runtime = runtime
         self._active_preset = active_preset
+        self._tracks_provider_generation = active_preset is None
         self._default_selection_signature = preset_helpers.default_selection_signature(
             runtime.snapshot_signature
         )

tests/agent/test_model_runtime_resolver.py

@@ -137,3 +137,52 @@ def test_resolver_refresh_preserves_unchanged_active_preset() -> None:
     assert refreshed is None
     assert resolver.runtime.provider is preset_provider
     assert resolver.model_preset == "fast"
+
+
+def test_resolver_refreshes_provider_generation_for_next_default_turn() -> None:
+    provider = _provider(temperature=0.2, max_tokens=2048)
+    resolver = ModelRuntimeResolver(_runtime(provider))
+    admitted = resolver.current()
+
+    provider.generation = GenerationSettings(temperature=0.8, max_tokens=512)
+    refreshed = resolver.current(refresh=True)
+
+    assert admitted.generation == GenerationSettings(0.2, 2048, None)
+    assert refreshed.generation == GenerationSettings(0.8, 512, None)
+
+
+def test_selected_preset_generation_does_not_fall_back_to_provider_defaults() -> None:
+    provider = _provider(temperature=0.1, max_tokens=1024)
+    resolver = ModelRuntimeResolver(
+        _runtime(provider),
+        model_presets={
+            "creative": ModelPresetConfig(
+                model="creative-model",
+                temperature=0.7,
+                max_tokens=4096,
+            )
+        },
+    )
+    selected = resolver.select_preset("creative")
+
+    provider.generation = GenerationSettings(temperature=0.9, max_tokens=64)
+    refreshed = resolver.current(refresh=True)
+
+    assert refreshed is selected
+    assert refreshed.generation == GenerationSettings(0.7, 4096, None)
+
+
+def test_resolver_mutates_only_its_default_selection() -> None:
+    initial = _runtime()
+    resolver = ModelRuntimeResolver(initial)
+
+    selected_model = resolver.select_model("next-model")
+    selected_window = resolver.select_context_window(65_536)
+
+    assert selected_model.model == "next-model"
+    assert selected_window.model == "next-model"
+    assert selected_window.context_window_tokens == 65_536
+    assert resolver.runtime is selected_window
+    assert resolver.model_preset is None
+    assert initial.model == "base-model"
+    assert initial.context_window_tokens == 10_000