[Feature Request] Add user guidance for visual/OCR tasks to reduce unnecessary processing time

Question

**Bug Description**
Tell Claude to ask user assisatance when it comes to visual tasks like this, it was extremley stupid by claude to try to understand a visual text, when user could read it in 10 seconds, claude spent like 40 mins.

**Environment Info**
- Platform: darwin
- Terminal: Apple_Terminal
- Version: 2.1.150
- Feedback ID: 98c3eee2-6bef-4798-9d64-e98dcfbfd8bc

**Errors**
```json
[]
```