fix: make model fallback chains configurable (#53)

Move hardcoded model fallback lists from module-level constants into
settings.fallback_models and settings.vision_fallback_models (pydantic
Settings fields). Can now be overridden via env vars
FALLBACK_MODELS / VISION_FALLBACK_MODELS or config/providers.yaml.

Removed:
- OLLAMA_MODEL_PRIMARY / OLLAMA_MODEL_FALLBACK from config.py
- DEFAULT_MODEL_FALLBACKS / VISION_MODEL_FALLBACKS from agent.py

get_effective_ollama_model() and _resolve_model_with_fallback() now
walk the configurable chains instead of hardcoded constants.

5 new tests guard the configurable behavior and prevent regression
to hardcoded constants.

tests/e2e/test_ollama_integration.py

@@ -33,7 +33,8 @@ async def test_ollama_connection():
 @pytest.mark.asyncio
 async def test_model_fallback_chain():
     """Test that the model fallback chain works correctly."""
-    from timmy.agent import DEFAULT_MODEL_FALLBACKS, _resolve_model_with_fallback
+    from config import settings
+    from timmy.agent import _resolve_model_with_fallback
 
     # Test with a non-existent model
     model, is_fallback = _resolve_model_with_fallback(
@@ -46,7 +47,7 @@ async def test_model_fallback_chain():
     # or the last resort (the requested model itself if nothing else is available).
     # In tests, if no models are available in the mock environment, it might return the requested model.
     if is_fallback:
-        assert model in DEFAULT_MODEL_FALLBACKS
+        assert model in settings.fallback_models
     else:
         # If no fallbacks were available, it returns the requested model as last resort
         assert model == "nonexistent-model"