How do I input images for inference or recognition?

[frozen30]

Feature description

Only text can be input through the main.py. Is there any way to input images for interaction?

Your Feature

No response

[1066196847]

这个无法直接实现，你看现在的主入口文件：run_mcp.py的run_default()函数，只有一个输入入口：prompt = input("Enter your prompt: ")，这里是输入文本的方式，而不是图片，你必须改造这里才行

[github-actions]

This issue has been inactive for 30 days. Please comment if you have updates.

[github-actions]

This issue was closed due to 45 days of inactivity. Reopen if still relevant.