Refactor LLM instantiation across agents and orchestrator

- Replaced direct instantiation of ChatOpenAI with a centralized get_llm function in CheckpointAgent, NoteAgent, ProjectAgent, and TaskAgent.
- Introduced a new llm.py module to handle LLM model instantiation and API key management.
- Updated settings.py to include LLM_MODEL and LLM_ROUTER_MODEL configurations.
- Modified orchestrator.py to use get_router_llm for intent classification.
- Updated requirements.txt to include litellm for LLM management.
- Adjusted tests to mock get_llm instead of ChatOpenAI directly.

app/core/orchestrator.py

@@ -6,10 +6,9 @@ import json
 from typing import Any, AsyncGenerator
 
 from langchain_core.messages import HumanMessage, SystemMessage
-from langchain_openai import ChatOpenAI
 
-from app.config.settings import settings
 from app.core.agent_registry import AgentRegistry
+from app.core.llm import get_router_llm
 from app.core.agent_registry import registry as _default_registry
 from app.schemas import ChatRequest, ChatResponse, ExecutionPlan
 
@@ -29,8 +28,8 @@ _SYNTHESIZE_HUMAN = (
 )
 
 
-def _make_llm(model: str = "gpt-4o-mini") -> ChatOpenAI:
-    return ChatOpenAI(model=model, temperature=0, api_key=settings.OPENAI_API_KEY)
+def _make_llm():
+    return get_router_llm()
 
 
 async def classify_intent(