Skip to content

Latest commit

 

History

History
 
 

plugin-builder

README.md

G-Assist Plugin Builder

Transform your ideas into functional G-Assist plugins with minimal coding! This tool uses OpenAI's Custom GPT to generate plugin code, making it easier than ever to extend G-Assist's capabilities. Whether you want to create a weather plugin, a task manager, or any other custom functionality, the Plugin Builder streamlines the entire development process.

What Can It Do?

  • Generate complete plugin code using AI
  • Create plugins that handle various types of queries
  • Works with the G-Assist plugin system
  • Minimal manual coding required

Before You Start

Make sure you have:

💡 Tip: Familiarize yourself with G-Assist's plugin architecture before starting your development journey!

Getting Started

GPT Plugin Builder

Plugin Builder GIF

Step 1: Generate Your Plugin

  1. Access the OpenAI Custom GPT for plugin generation
  2. Start by clearly describing your plugin's purpose and functionality. For example:
    • "I want to create a weather plugin that can fetch current weather for any city"
    • "I need a plugin that can manage a todo list with voice commands"
  3. The GPT will guide you through the process and generate:
    • Main plugin code
    • Manifest file
    • Configuration file (if needed)
    • Installation instructions

💡 Tip: For best results, after the files are generated, ask the GPT to "Make sure the files follow the G-Assist Plugin template specifications". This helps catch and correct any potential inconsistencies.

Step 2: Review and Customize

  1. Examine the generated code:

    • Check that all required imports are present
    • Ensure all handler functions are properly defined
    • Review any configuration parameters
    • Ensure file paths for input and output files are correct

  2. Make necessary adjustments:

    • Add error handling where needed
    • Customize response messages
    • Adjust command triggers to your preference
    • Update any hardcoded values

  3. Test the plugin functionality:

    • Test each command individually
    • Verify error cases are handled gracefully
    • Check that responses are formatted correctly
    • Ensure the plugin integrates properly with G-Assist

💡 Tip: Ask the GPT to add logging statements to help with debugging. Example:

import logging
import os

LOG_FILE = os.path.join(os.environ.get("USERPROFILE", "."), 'your_plugin.log')
logging.basicConfig(filename=LOG_FILE, level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")

logging.info('G-Assist plugin initialized.')

💡 Tip: Keep the manifest.json, plugin.py and build.bat files in sync when making changes to command handlers or plugin names.

Step 3: Build and Deploy

Follow the standard G-Assist plugin deployment process:

pyinstaller --onedir --name your-plugin-name your_plugin.py

Step 4: Install the Plugin

  1. Create a new folder in the G-Assist plugins directory: 
    %programdata%\NVIDIA Corporation\nvtopps\rise\plugins\your-plugin-name
  2. Copy your built plugin files to this directory
  • your-plugin-name.exe
  • config.json
  • manifest.json

Example Use Cases

You can create plugins for:

  • Weather information
  • Task management
  • Calendar integration
  • Smart home control
  • Custom data fetching

💡 Tip: Start with a simple plugin to understand the workflow before tackling more complex projects!

Troubleshooting Tips

  • Generation issues? Make sure your plugin description is clear and specific
  • Build errors? Verify all dependencies are properly installed
  • Plugin not working? Double-check the deployment folder structure

Want to Contribute?

We welcome contributions to improve the Plugin Builder! Whether it's:

  • Adding new templates
  • Improving code generation
  • Enhancing documentation
  • Fixing bugs

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Need Help?

If you run into any issues:

  1. Check the troubleshooting section above
  2. Review the generated code for any obvious issues
  3. Verify your G-Assist installation
  4. Make sure your OpenAI API access is working correctly

Acknowledgments

Special thanks to:

  • OpenAI for their Custom GPT technology
  • All contributors who help improve this tool