Upgrade to GPT-4o and improve documentation

[lancewillett]

Thanks for this amazing tool! I've set it up in a Dropbox folder I have on macOS to process new screenshots

Summary

• Upgrade from gpt-4-vision-preview to gpt-4o with proper API format
• Remove PIL dependency (gpt-4o handles full resolution images)
• Use spaces in filenames instead of underscores for better readability
• Fix documentation typos and add troubleshooting tips

Suggested changes

Model upgrade:

• Switch to gpt-4o
• Fix image_url format to match current OpenAI API spec (object with url key)
• Remove image resizing—gpt-4o handles full resolution

Better filenames:

• Spaces instead of underscores
• Sentence case
• Up to 10 words for more descriptive names

Documentation:

• Fix "Automater" typo
• Add note about .bashrc for non-zsh users
• Add troubleshooting for Full Disk Access and Folder Actions Setup
• Update rate limit note to be model-agnostic

Test plan

• Take a screenshot and verify it gets renamed with spaces and sentence case
• Confirm gpt-4o API calls work without PIL installed