Screenshot Reader
Analyze screen content using Vision AI.
When to use
When user says what's on screen, read this, OCR, or analyze image.
How to use
- analyze_image (auto-captures screen if no image provided)
- Parse the Vision AI response
- Present extracted text or image description
- If user asks follow-up: use the extracted content as context
Example
User: "屏幕上显示了什么？" → analyze_image (auto-captures screen) → "屏幕上显示的是时钟界面，当前时间 15:30，日期 3月28日。"