BetterIDEa, a web-based Integrated Development Environment (IDE) for building Lua processes on AO, is seeking to enhance its functionality with AI-powered code suggestions.

The primary goal of this bounty is to implement AI code suggestions using Large Language Models (LLMs) directly in the IDEs editor. This would leverage the power of LLMs to provide context-aware code completions, improving developer productivity and code quality for lua processes.

An ideal solution will seamlessly integrate with the existing BetterIDEa user interface (.lua files, lua notebooks & markdown editors), providing real-time code suggestions as developers type. It should be able to understand the context of the Lua code being written for AO processes and offer relevant, helpful completions that align with best practices and common patterns in Lua development.

Submissions will be judged on factors such as accuracy of suggestions, response time, user experience, and overall integration with the BetterIDEa platform. The winning solution should enhance the development workflow without disrupting the existing features of the IDE.

(For reference and examples checkout VSCode's github copilot extension, See here for an extension worth checking out to get VSCode copilot like autocomplete.)

Steps to take on this challenge:

1. Review the video guide created by BetterIDEa's Ankush Kun

   
   

2. Fork BetterIDEa

   
   

3. Implement your changes: Particularly in the notebook-editor.tsx and single-file-editor.tsx files

   
   

4. A early stage base for the ai autocomplete function has already been added through npms @ai-sdk/google and Google Gemini services & some @lib/ai-* files. It is recommended to build and enhance this implementation, but if you prefer, you are welcome to use your own approach.

   
   

5. If using an off chain service for ai autocomplete, Github Copilot is recommended.

Prizing:

This bounty will be selected by judges from the BetterIDEa team, who will award $1500 USD in AR total to a single winner.