Screenshots

Screenshot Intelligence

Clipboard Recast includes powerful screenshot intelligence features that allow you to analyze, extract text from, and understand screenshots using vision-capable AI models.

How Screenshot Detection Works

macOS provides two different ways to take screenshots, and Clipboard Recast supports both:

Method 1: Screenshots Saved to Desktop (Recommended)

Keyboard Shortcuts:

• Cmd+Shift+3 - Full screen screenshot → Saved to Desktop
• Cmd+Shift+4 - Selection screenshot → Saved to Desktop

How Clipboard Recast Handles This:

• ✅ Automatically detected - The app monitors your Desktop for new screenshot files
• ✅ Automatically added to clipboard history within 1-2 seconds
• ✅ Shows visual thumbnail in history window (if available)
• ✅ Triggers screenshot workflows if configured

File Naming Pattern: Clipboard Recast detects these standard macOS screenshot filenames:

• Screenshot 2026-01-14 at 7.44.01 PM.png
• Screen Shot 2026-01-14 at 7.44.01 PM.png

Method 2: Screenshots Copied to Clipboard

Keyboard Shortcuts:

• Cmd+Shift+Control+3 - Full screen → Copied to clipboard
• Cmd+Shift+Control+4 - Selection → Copied to clipboard

How Clipboard Recast Handles This:

• ✅ Detected via clipboard monitoring
• ✅ Immediately added to history
• ✅ No file saved to Desktop
• ✅ Triggers screenshot workflows

Method 3: Copy Image from Applications

You can also copy images directly:

• Right-click on image → “Copy Image”
• Press Cmd+C on selected image in any app
• ✅ Detected via clipboard monitoring

Using Screenshot Features

1. View Screenshot in History

1. Take a screenshot using any method above
2. Wait 1-2 seconds for detection
3. Click menu bar icon → History (or press Cmd+H)
4. See your screenshot with visual thumbnail

2. Analyze Screenshots with AI

Transform screenshots just like text:

1. Take a screenshot
2. Press Cmd+Shift+A (or click menu bar → Clipboard Actions)
3. Choose an action:
   • Explain - Describe what’s in the screenshot
   • Summarize - Summarize the visible content
   • Custom - Ask a specific question about the image

3. Create Automatic Screenshot Workflows

Set up workflows to automatically analyze screenshots:

1. Click menu bar icon → Workflows (or press Cmd+R)
2. Click Add Workflow
3. Configure:
   • Name: “Auto-analyze Screenshots”
   • Trigger Type: Screenshot
   • Action: Explain (or any other action)
   • Auto-trigger: ✓ (enable for automatic processing)
4. Click Save

Now every screenshot you take will be automatically analyzed!

4. Search Screenshot History

If history is available:

1. Open History window
2. Type in search field
3. Screenshots are searchable by their AI analysis
4. Double-click to copy back to clipboard

Use Cases

1. Error Debugging

Scenario: Error appears on screen that you can’t easily copy

Workflow:

1. Take screenshot of the error (Cmd+Shift+4)
2. App auto-analyzes (if workflow enabled)
3. Get debugging suggestions and solutions
4. Copy solution and apply fix

Alternative: Manual process

1. Take screenshot
2. Press Cmd+Shift+A
3. Select “Search Solutions”
4. Review solutions

2. UI/UX Feedback

Scenario: Reviewing a design or interface

Workflow:

1. Screenshot the design
2. Press Cmd+Shift+A
3. Select “Custom” action
4. Ask: “What could be improved about this UI?”
5. Get actionable feedback

3. Code Review from Screen

Scenario: Code visible on screen but not easily copyable

Workflow:

1. Screenshot the code
2. Press Cmd+Shift+A
3. Select “Explain”
4. Get code explanation with potential improvements

4. OCR Text Extraction

Scenario: Text in image or PDF that can’t be selected

Workflow:

1. Screenshot the text
2. Press Cmd+Shift+A
3. Select “Custom” action
4. Prompt: “Extract all text from this image”
5. Copy extracted text

5. Diagram Understanding

Scenario: Understanding flowcharts, architecture diagrams, or charts

Workflow:

1. Screenshot the diagram
2. Press Cmd+Shift+A
3. Select “Explain”
4. Get step-by-step explanation of the diagram

6. Documentation from Screenshots

Scenario: Creating documentation from UI screenshots

Workflow:

1. Take screenshots of each step
2. For each screenshot:
   • Press Cmd+Shift+A
   • Select “Explain”
   • Copy explanation
3. Paste explanations into documentation

Supported AI Providers

Not all AI providers support vision/image analysis. Here are the ones that do:

✅ Claude (Anthropic)

• Recommended for screenshot analysis
• Excellent vision capabilities
• Accurate text extraction
• Good at understanding UI and diagrams

✅ OpenAI (GPT-4 Vision)

• Great for detailed analysis
• Strong OCR capabilities
• Good at understanding complex images

✅ Google Gemini

• Good for charts and diagrams
• Strong multimodal understanding
• Decent OCR capabilities

✅ OpenRouter

• Access to multiple vision models
• Can choose best model for task
• Variety of pricing options

❌ Non-Vision Providers

These providers don’t support image analysis:

• Cerebras
• Together AI (most models)
• z.ai

If you try to analyze a screenshot with a non-vision provider, you’ll receive an error message.

Best Practices

1. Use Desktop Screenshots for Auto-Detection

Recommended: Cmd+Shift+4 (saves to Desktop)

• More reliable detection
• Creates permanent file you can reference
• Triggers auto-workflows more consistently

Less Reliable: Cmd+Shift+Control+4 (clipboard only)

• Works but requires clipboard monitoring
• No permanent file created
• May miss if clipboard changes quickly

2. Wait for Detection

After taking a screenshot:

• Wait 1-2 seconds for detection
• Screenshot monitoring checks every ~1 second
• Don’t copy other content immediately after

3. Take Clear Screenshots

For best AI analysis results:

• Ensure good lighting/contrast
• Capture relevant area only
• Avoid very small text
• Use higher resolution when possible

4. Choose the Right AI Provider

Different providers excel at different tasks:

• Code in screenshots: Claude or OpenAI
• UI/Design analysis: Claude
• Charts/Diagrams: Gemini or Claude
• Text extraction (OCR): OpenAI or Claude

5. Be Mindful of Privacy

Screenshots may contain sensitive information:

• Disable auto-trigger for screenshot workflows if you frequently screenshot sensitive data
• Remember that screenshots are sent to your AI provider
• Screenshots are stored locally in history

6. Use Specific Prompts

For custom actions, be specific:

• ❌ “Analyze this”
• ✅ “Extract all text from this image in order”
• ✅ “List all UI elements and their purposes”
• ✅ “What error is shown and how do I fix it?”

Tips and Tricks

Combine with Workflows

Create specialized screenshot workflows:

“Extract Code from Screen”

• Trigger: Screenshot
• Action: Custom
• Prompt: “Extract all code from this screenshot with proper formatting”
• Auto-trigger: ✗ (manual)

“Analyze Error Screenshots”

• Trigger: Screenshot
• Action: Search Solutions
• Auto-trigger: ✓ (automatic)

“Document UI Flow”

• Trigger: Screenshot
• Action: Custom
• Prompt: “Describe this UI screen for documentation purposes”
• Auto-trigger: ✗ (manual)

Quick Screenshot Analysis Pattern

The fastest way to analyze screenshots:

1. Position what you want to capture
2. Cmd+Shift+4 → Select area
3. Wait 2 seconds
4. Cmd+Shift+A → Select action
5. Review result

With auto-trigger workflow:

1. Cmd+Shift+4 → Select area
2. Wait 2-3 seconds
3. Result appears automatically

History Management

Screenshots can fill up history quickly:

• Keep auto-trigger selective
• Manually delete screenshot history items
• Consider disabling screenshot auto-detection if not needed

Testing Screenshot Workflows

Before enabling auto-trigger:

1. Create workflow with auto-trigger disabled
2. Take 3-5 test screenshots
3. Manually analyze each
4. Verify results are useful
5. Then enable auto-trigger

Troubleshooting

Screenshots Not Appearing in History

Possible causes:

• Using clipboard-only shortcuts (Cmd+Shift+Control+4)
• Screenshots not saving to Desktop
• Non-standard screenshot naming
• Wait longer (2-3 seconds)

Solutions:

• Use Cmd+Shift+4 instead (saves to Desktop)
• Check Desktop for screenshot files
• Verify filename matches pattern

AI Not Analyzing Screenshots

Possible causes:

• AI provider doesn’t support vision
• API key not configured
• Network connection issue
• Screenshot too large or complex

Solutions:

• Switch to Claude, OpenAI, or Gemini
• Verify API key in Settings
• Check internet connection
• Try smaller/clearer screenshot

Auto-Trigger Not Working

Possible causes:

• “Auto-process clipboard changes” disabled
• Workflow auto-trigger not enabled
• Screenshot not detected yet

Solutions:

• Enable master auto-process switch
• Enable workflow’s auto-trigger
• Wait 2-3 seconds after screenshot

Poor Analysis Quality

Possible causes:

• Screenshot unclear or low quality
• Wrong AI provider for task
• Ambiguous prompt

Solutions:

• Take clearer screenshot
• Try different AI provider
• Use more specific custom prompt

Next Steps

• Vision Models: Learn about vision model capabilities
• Workflows: Create advanced screenshot workflows
• Usage Guide: Integrate screenshots into your workflow