- 🔭 I'm currently working on: Agent · GUI Agent · Multimodal Understanding (VLM) · Omni / Unified Model · Code-driven Generation · AIGC
- 🌱 I'm currently learning: Agentic systems / Harness & unified multimodal models & code generation (HTML / CSS / SVG / LaTeX) · 📈 stock investing
- 📫 How to reach me:
- 📄 CV / Resume:
- 🛠️ Skills:
- 🤖 AI Coding:
- CMA-Harness — Cognitive-structured Multimodal Agent Harness: memory-centric agent for long-horizon multimodal understanding, generation & editing (with M2CA-Bench)
- MLNLP-World/Top-AI-Conferences-Paper-with-Code — contributor
- Kimi Work Pets — Desktop pets for the Kimi Work app: one-click install of 14 Codex Pets community sprites (Dimo, Hiyuki, Hu Tao, Miyabi, Violet...), with hover interactions & gaze-following head tracking (macOS / Windows)
- Cadence — CPU-scheduling-inspired task manager for multi-task life: suspend tasks + adaptive backoff polling reminders
- Stock-assistant — US equities reference dashboard: watchlist, multi-timeframe charts, composite signals (FastAPI + Next.js + Alpaca)
- YouTubeMusic-MacOS — Native macOS desktop client for YouTube Music with media keys, tray & Control Center integration
- Mymd-editor — Typora-style Markdown editor for Windows (Electron + Milkdown)
