[Feature]: add litellm for llm completion and tool-call

[gusye1234]

Feature Description

Currently, acontext-core uses self-written llm abstract to support OpenAI and Anthropic SDK, which is not necessary, easy to have bugs, and may not support some latest models (gpt-5.2, gemini...).

We plan to use litellm to proxy llm SDK.

Impact Component/Area

Which part of Acontext would this feature affect?

• Client SDK (Python)
• Client SDK (TypeScript)
• Core Service (Python)
• API Server (Go)
• UI/Dashboard (Next.js)
• CLI Tool
• Documentation
• Other (please specify)

[AutoLTX]

Has a feeling this one should be prioritize, as claude skills native support(#95) needs to define several functions for scenarios like executing bash. If litellm already adopted, we can define and pass litellm type function param directly. Otherwise will have additional migration cost.