Photo Clipper

Command Line

python photo_clipper.py <photo_path> "<prompt>" [output_path]

Examples

Remove empty sky:

Remove 30% from top to reduce empty sky

Focus on subject:

Crop to focus on the person's face in the center

Remove distractions:

Remove the partial tree on the left edge

Improve composition:

Apply rule of thirds to balance the composition

Trim specific amount:

Remove 10% from all edges for a cleaner frame

Output

• Returns path to clipped photo
• Never overwrites original
• Format: {original_name}-clipped.{ext}
• Displays GPT-5's confidence score and reasoning

Error Handling

• Falls back to safe defaults (no clipping) if GPT-5 unavailable
• Validates parameters (max 50% removal per dimension)
• Clear error messages for missing files or invalid formats

Requirements

• OPENROUTER_API_KEY environment variable must be set
• Photo must be JPG or PNG format
• Photo under 50MB recommended

Technical Implementation

See references/implementation.md for:

• GPT-5 vision API integration
• Cropping algorithm details
• Error handling strategy
• Performance optimization

Examples

See examples/common-use-cases.md for:

• Remove empty sky
• Focus on subject
• Remove distractions
• Improve composition
• Error handling scenarios