Execution failed: CAPIError: 400 model "claude-opus-4.6" not supported for vision

### Describe the bug

Execution failed: CAPIError: 400 model "claude-opus-4.6" not supported for vision getting this issue for all teh prompts , how to make thsi session to come out of this issue

### Affected version

_No response_

### Steps to reproduce the behavior

In github copilot cli ,

- Include an image content block type.
- Ask coplot to review teh image

### Expected behavior

Getting this Execution failed: CAPIError: 400 model "claude-opus-4.6" not supported for vision

### Additional context

Unable to get rid of this issue and tried to restart teh sessiion , chnage the model to claude opus , gpt but results is same

首先，使用 image content block 和受影響的模型選擇，重現 GitHub Copilot CLI 中的影像審查流程。追蹤工作階段如何選擇 vision model，並驗證 image prompts 要嘛使用受支援的模型，要嘛回傳明確的復原路徑，而不是 400 錯誤。

由索引模型根據 Issue 內容生成。