feat(provider): add Huawei ModelArts provider support

[Ferris0518]

Summary

• add Huawei ModelArts as a supported LLM provider
• wire the new adapter into the default model adapter manager and CLI doctor checks
• add a sample client config, documentation updates, and unit tests
• remove an incorrect README description to keep provider documentation accurate

What changed

• implemented a new huawei-modelarts model adapter using an OpenAI SDK-compatible chat completions interface
• registered the adapter in the model adapter exports and default adapter manager
• updated CLI adapter help text and doctor diagnostics to recognize huawei-modelarts
• added client/examples/config.huawei-modelarts.example.json
• added unit tests for adapter loading, doctor report validation, and ModelArts adapter behavior
• corrected inaccurate README content related to provider support and configuration

Why

This PR adds first-class support for Huawei ModelArts so users can configure and run the framework with that provider
using the same adapter-based workflow as existing providers. It also ensures the README stays aligned with the actual
supported integrations.

Testing

• added unit tests for the Huawei ModelArts adapter
• added coverage for adapter manager loading
• added coverage for CLI doctor validation
• verified documentation and sample config updates