Introducing the AI-Powered Debugger for Music Blocks – GSoC 2025

Hi everyone! I'm Sneha Poojary, a GSoC 2025 aspirant working with Sugar Labs. I’m thrilled to share my journey as I develop an AI-Powered Debugger for Music Blocks.

This project aims to bridge the gap between users’ creative intentions and the technical challenges they face while composing with Music Blocks. By integrating a lightweight AI model, the debugger will assist users in understanding, troubleshooting, and improving their block-based musical code — all in real time.

💡 Why This Project?

Music Blocks is a visual programming language for exploring musical concepts. While it's intuitive, users (especially beginners) may struggle when their musical programs don't work as expected.

My project introduces a natural language-based debugging assistant that can:

  • Explain what a user’s block arrangement is doing.
  • Identify logical errors or unintentional musical outputs.
  • Suggest fixes or improvements.
  • Respond to queries like: “Why doesn’t this play the rhythm I intended?” or “How can I make this repeat 4 times instead?”

🛠️ Tech Stack

Here’s a quick look at the tools and technologies I’ll be using:

  • Frontend: JavaScript/HTML (integrated into Music Blocks)
  • Backend: FastAPI (Python)
  • AI Model: Lightweight LLM with Retrieval-Augmented Generation (RAG)
  • Storage: JSON-based project analysis and context mapping

📌 Current Status

  • ✅ Submitted proposal: GSoC 2025 Proposal – AI-Powered Debugger (Link)
  • 🔍 Exploring Music Blocks codebase and existing debugging flow
  • 📚 Reading documentation, tutorials, and getting feedback from mentors
  • 🤝 Connecting with the Sugar Labs community on Matrix and GitHub

🚀 Next Steps

  • Finalize development plan during the Community Bonding Period
  • Create mock UI for chatbot interactions within the Music Blocks interface
  • Start modular backend development (project analysis > issue classification > response generation)

🙌 Stay Tuned

I'll be posting weekly updates throughout GSoC 2025 under this blog series. If you’re interested in open source, educational tools, music, or AI – this journey will definitely strike a chord! 🎶

Feel free to drop your thoughts or questions in the comments 💬

Thank you Sugar Labs and Google for this incredible opportunity 💜

Let’s build something amazing together!

— Sneha ✨