This VSCode extension demonstrates using Large Language Models (LLMs) for active program debugging by providing real-time runtime context. It's designed for developers and researchers seeking to enhance debugging efficiency and explore novel AI-assisted development workflows.

How It Works

The extension integrates with Node.js debug sessions, acting as a debug adapter tracker. It captures runtime information like variable values, function calls, and branch decisions, feeding this live data alongside static code to an LLM. The LLM then analyzes this enriched context to suggest debugging actions, such as setting breakpoints or stepping through code, which the extension executes automatically.

Quick Start & Requirements

• Install via VSIX: npm run build in the project root, then use VSCode's "Extensions: Install from VSIX..." command.
• Development: Open project in VSCode and run "Debug: Start Debugging".
• Requires Node.js debugging sessions.
• See GitHub repository for details.

Highlighted Details

• Integrates live debugging information (variables, stack traces, breakpoints) into LLM context.
• Automatically sets initial breakpoints based on code analysis and LLM recommendations.
• Monitors runtime events (exceptions, thread stops) for detailed state gathering.
• Supports common debug operations: step over, step in, step out, continue.
• Generates synthetic runtime data beyond static code analysis.

Maintenance & Community

This project is a proof of concept and research experiment; it will not be actively maintained or further developed.

Licensing & Compatibility

The license is not explicitly stated in the README. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The project is experimental and will not be actively maintained or further developed, indicating potential for unaddressed bugs or lack of future feature support.