← back to blog

Why we built yoinko

Every notes app we tried fell into one of two traps: either it locked our files into a proprietary format we couldn't read without the app, or it bolted AI on as an afterthought — a chatbot floating in the corner with no real connection to what we were writing.

We wanted something different. An editor where the AI actually knows the page you're working on. Where it can draft a section, rewrite a paragraph, or summarize the whole thing — and you can apply the result with one click. But also an editor where your files are just files. Markdown on disk. A SQLite database you can open with any tool.

So we built yoinko.

The architecture is intentionally simple: a Node.js server, a vanilla TypeScript client bundled with esbuild, and SQLite for everything. No framework on the frontend, no ORM on the backend. The AI layer is a thin proxy — your API key goes straight to your chosen provider (OpenAI, Gemini, Claude, Ollama, whatever). We never see it, never store it on our servers.

We made it open source from day one because we believe the best tools are the ones you can inspect, modify, and own. If yoinko disappears tomorrow, your data doesn't. It's right there in the data/ folder.

We're just getting started. GitHub sync, a tasks page, mobile readers, and a plugin API are all on the roadmap. But the core philosophy won't change: your notes, your machine, your rules.