Digital Tech Director at Starbucks China. I build AI agents that connect the physical retail world to the AI ecosystem.
By day I lead digital technology for one of the largest coffee chains in the world. By night I ship open-source tools that make AI agents actually useful for businesses.
| Project | What it does |
|---|---|
| coffee-mcp | Multi-brand MCP platform for coffee & tea ordering — connect any beverage brand to any AI assistant in 2-3 weeks |
| SkillEngine | Framework-agnostic skills engine for LLM agents — Claude Code-like experience on any model |
| claude-channel-weixin | WeChat channel for Claude Code — chat with Claude in WeChat |
| sawzhang-skills | Claude Code skills collection — CCA study guides, Twitter research, MCP review tools |
- AI + Retail — How AI agents become the next ordering channel for consumer brands
- MCP (Model Context Protocol) — The open standard connecting AI assistants to real-world tools and data
- Agent Skills — Portable, framework-agnostic skill systems that work across any LLM provider
- Practical AI — Not demos. Shipped systems that handle real orders, real users, real money
Python TypeScript MCP Claude Code OpenAI LLM Agents Retail Tech
- Twitter/X: @chenhua_zhang
- ClawHub: @sawzhang