DEV Community

D4vRAM
D4vRAM

Posted on

I built Debuggle — a desktop app that explains any error at 3 levels of depth

Every developer has been there: you paste an error into Google, find a Stack Overflow answer from 2014, and still have no idea what actually went wrong.

I built Debuggle because I was that developer — and I still am, honestly.

Debuggle — Your personal debugging assistant. Analyze · Learn · Save

I've only been coding since late 2024. I needed a tool that didn't assume I already knew what a null pointer exception meant at a deeper level, or why a specific Node.js error was happening in my context, not just generically.

So I built it.

What Debuggle does

Debuggle app interface showing the Analyze tab with Beginner, Medium and Expert level selector

You paste any stack trace or error message. Debuggle analyzes it and gives you three explanations simultaneously:

  • Novice — plain language, no jargon, just "here's what broke and why"
  • Mid — technical context, the cause, what to look for
  • Expert — root cause, code-level diagnosis, fix suggestions

After the analysis, you can open a multi-turn chat about that specific error with the context already loaded. You're not starting from scratch explaining the situation — Debuggle already knows what you're dealing with.

It also saves analyses to a personal vault, so over time you build a searchable knowledge base of every error you've ever debugged.

The privacy part

I care about this, so I built it in from the start.

Debuggle supports 6 AI providers: Groq, OpenRouter, Ollama, Anthropic, OpenAI, and VeniceAI. That last one and Ollama mean you can run it 100% locally — no data leaves your machine.

API keys are stored in your OS keychain. Never written to disk in plaintext.

The stack

Electron 33 + React 18 + TypeScript. Cross-platform: Windows, Linux, macOS.

I'm self-taught and about 8 months into building real projects. This is probably the most complete thing I've shipped so far — multi-provider AI support, a vault system, pattern tracking, context-aware chat. It took longer than I expected and I learned a lot building it.

Why I'm sharing this

Partly to get feedback. Partly because I think there's a gap between "paste your error into ChatGPT" and having a tool that's actually built around the debugging workflow — with context persistence, level-appropriate explanations, and local-first privacy.

If you're a developer who's ever felt like error messages are written for someone else, Debuggle might be for you.

GitHub: github.com/D4vRAM369/Debuggle

Would love to hear what you think — or what you'd add.
What's the most frustrating type of error to debug? 👇

Top comments (0)