feat: add qwen3-vl-plus vision model support for Qwen Code CLI

[roomote]

Summary

This PR attempts to address Issue #8993 by adding support for the qwen3-vl-plus vision model, enabling users of Qwen Code CLI to access this vision-language model through Roo Code.

Changes

• Added qwen3-vl-plus to the QwenCodeModelId type union
• Added model configuration for qwen3-vl-plus with:
  • Vision support enabled (supportsImages: true)
  • 1M context window (matching other Qwen3 models)
  • 65K max tokens (matching other Qwen3 models)
  • Appropriate description for the vision-language model

Implementation Notes

The model parameters (maxTokens: 65_536, contextWindow: 1_000_000) were set to match the existing qwen3-coder-plus and qwen3-coder-flash models for consistency. If different values are needed based on official Qwen specifications, these can be adjusted.

Testing

• ✅ All existing tests pass
• ✅ TypeScript compilation successful
• ✅ Build completes without errors
• ✅ Linting passes

Related Issues

Closes #8993

Feedback

Feedback and guidance are welcome, especially regarding the specific model parameters if they should differ from the other Qwen3 models.

---

Important

Add support for qwen3-vl-plus vision-language model to Qwen Code CLI with specific configuration in qwen-code.ts.

• Model Support:
  • Add qwen3-vl-plus to QwenCodeModelId in qwen-code.ts.
  • Configure qwen3-vl-plus with supportsImages: true, 1M context window, and 65K max tokens.
• Testing:
  • All existing tests pass.
  • TypeScript compilation and build complete without errors.
  • Linting passes.

^{This description was created by for 00f20de. You can customize this summary. It will automatically update as commits are pushed.}

[roomote]

Code review complete. No issues found - the implementation correctly adds qwen3-vl-plus model support with appropriate vision capabilities enabled.

_{Mention @roomote in a comment to trigger your PR Fixer agent and make changes to this pull request.}