feat: add Novita AI as optional LLM provider

[Alex-wuhu]

This PR adds Novita AI as an optional LLM provider.

Changes

• Added NovitaProvider implementation following the existing provider pattern
• Registered Novita in the provider manager
• Supports DeepSeek-V3, DeepSeek-R1, Llama 3.3, Llama 3.2, Gemma 2, Mistral Nemo, and Qwen 2.5 models
• Uses OpenAI-compatible endpoint: https://api.novita.ai/openai

Testing

• Follows existing provider architecture pattern
• No breaking changes to existing functionality