I care about where software meets judgment — the thin layer where agents, products, and humans actually collaborate.
Most of my work lives in three currents:
Day job is large-scale product engineering at ByteDance. Public writing and open source stay intentionally abstract — the interesting details are usually the ones that shouldn't leave the building.
|
Minimal macOS translate-on-select. Swift · AppKit · Vision · system Translation |
Self-hosted multi-user local AI agent system. Chat, tool traces, workspace ops — an attempt to make agent teams feel like real software, not a terminal demo. TypeScript · Node · agent orchestration |
|
A small Chrome experiment for posture-aware sitting. Local camera, local storage — proof that useful software can stay on-device. Chrome extension · on-device ML UX |
Older notes from the climb: frontend fundamentals, systems thinking, and the habit of writing things down. learning log · 日拱一卒 |
What I'm exploring (public-safe snapshot)
From recent local craft — without naming internal products:
- Coding agents & workbenches — orchestration, skills, eval harnesses, and desktop shells around Claude Code / Codex-class runtimes
- Visual feedback for agents — browser-side annotation that turns UI intent into structured agent input
- Native utilities — macOS tools that respect the OS instead of fighting it
- Small games & engines — canvas physics, browser 3D, games as a lab for deterministic simulation
- Language surface area — TypeScript for product velocity, Go for durable services
- Designing agent collaboration that feels like a team, not a chat box
- Collecting sharp tools at the edge of AI × systems × craft
- Open to thoughtful collaboration on open-source agent infrastructure and developer experience
Hangzhou · China · UTC+8 · Less noise. More signal.