feat: add LiteLLM as unified LLM provider

[RheagalFire]

Summary

Adds LiteLLM as a new LLM provider option (LLM_PROVIDER=litellm), giving QuantDinger users access to 100+ LLM providers (Anthropic, Azure, Bedrock, Ollama, Groq, Together, Fireworks, etc.) through litellm.completion() as an SDK dependency. Additive only - all existing providers untouched.

Changes

• backend_api_python/app/services/llm.py: added LITELLM = "litellm" to LLMProvider enum, _call_litellm() method with drop_params=True, dispatch in call_llm_api()
• backend_api_python/app/config/api_keys.py: added LITELLM_API_KEY property (optional)
• backend_api_python/app/routes/settings.py: added LiteLLM to provider dropdown and settings group (API Key, Model, Base URL)
• backend_api_python/requirements.txt: added litellm>=1.80,<1.87

Test plan

• Tested locally with docker compose up -d --build
• Backend logs show no errors
• Relevant pytest tests pass (691 passed, 5 skipped, 0 failed)

API documentation (if routes/schemas changed)

• No routes/schemas changed - LiteLLM added as backend provider option only
• No /api/agent/v1 routes changed
• No breaking API changes

Screenshots (if UI change)

NA

[RheagalFire]

cc @brokermr810