Show HN: Support Copilot

Every developer and support engineer knows this scenario: you're on a support call, the customer shares their screen showing an issue, and you need help figuring it out. Traditionally, this meant:

• Taking screenshots manually
• Pasting them into chat tools
• Asking colleagues or AI for help
• Repeating the process multiple times

This screenshot-paste-ask-repeat cycle is inefficient and breaks your flow during critical support situations.

Support Copilot streamlines this workflow by:

• Automatically capturing screenshots every 5 seconds during support calls
• Keeping screenshots local until you choose to share them
• Integrating with GitHub Copilot for instant AI assistance
• Maintaining context across multiple queries in the same session

1. Start a support call - Customer shares their screen
2. Launch Support Copilot - Begin automatic screenshot capture
3. When you need help - Select a predefined prompt template
4. Add context - Include the latest screenshot and additional details
5. Get AI assistance - GitHub Copilot analyzes the visual context and provides guidance
6. Iterate as needed - Continue the conversation with maintained context

Note: All screenshots remain on your local machine until you explicitly choose to include them in your AI queries.

• Python 3.10+ (currently tested on macOS only)
• Visual Studio Code (latest version)
• GitHub Copilot subscription with image-capable model
  • Claude Sonnet recommended for best results
  • Other vision-capable LLMs also supported

• ✅ Automatic screenshot capture (configurable interval)
• ✅ Local storage with privacy controls
• ✅ Predefined prompt templates
• ✅ Context persistence across queries
• ✅ GitHub Copilot integration
• ✅ Cross-platform support (macOS tested)

We welcome contributions! Here's how to get started:

1. Fork this repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes and test them
4. Submit a pull request with a clear description

Let's build the future of AI-assisted support together! 🚀

[Add your license here]

If you encounter issues or have questions:

• Open an issue on GitHub
• Check the documentation in /docs
• Review existing discussions