The Problem
When debugging Android apps on macOS, Logcat outputs thousands of lines per second. Finding the actual error in that wall of text is exhausting — especially for developers who aren't familiar with Android internals.
I wanted something smarter. Not just a viewer, but a tool that could say "here's what went wrong, and here's how to fix it."
What It Does
HiyokoLogcat has a one-click AI diagnosis feature. When you spot something suspicious in the log, you hit the diagnose button — and within seconds, Gemini Flash analyzes the context and returns a plain-English explanation with a suggested fix.
No copy-pasting into ChatGPT. No switching apps. Just instant answers inside the viewer itself.
How It Works (Without Giving Away Too Much 🐤)
The key insight was context. Sending just one error line to the AI gives you a mediocre answer. Sending the surrounding lines — what happened before and after — gives you something genuinely useful.
Before anything gets sent to the API, sensitive data like IP addresses and email addresses are masked locally. Your logs never leave your machine in raw form.
The Tricky Part
Getting the balance right between context size and API response time took some tuning. Too little context and the AI misses the cause. Too much and the response gets slow and expensive.
The other challenge was making it feel fast. Logcat moves quickly, and a sluggish AI response breaks the flow. Rust handling the log pipeline helped a lot here.
Try It Free
HiyokoLogcat is free and open source.
👉 GitHub: https://github.com/Hiyoyoko/HiyokoLogcat