Zephyr Gao

Cloud Native and Delivery Platform Engineer focused on Kubernetes controllers, durable workflow orchestration, and agent-first engineering.

I lead the technical architecture and implementation of a long-term convergence between online and offline delivery systems. I also build tools and repository workflows for AI-assisted software engineering.

Current Work

• Leading cross-system delivery-platform architecture and phased migration
• Designing durable workflow scheduling and immutable execution inputs
• Building agent-first repository workflows and quality gates
• Writing about systems, delivery platforms, and AI agents

Featured Projects

• skill-forge — reusable skills for agent documentation, review, verification, and quality gates
• opencode-plugin — tooling for installing, managing, and synchronizing Claude Code plugins in OpenCode
• Zephyr's Lab — notes on agents, kernels, delivery platforms, and distributed systems

Selected Writing

• Repository as an Agent Environment
• Staff Is Not Team Size
• Delivery Platform Concurrency Patterns

Core Stack

Go · Kubernetes · controller-runtime · client-go · Argo Workflows · KubeVela · Helm · Docker · Python

Interests

AI Agents · Software Architecture · Distributed Systems

Education

Peking University — B.S. in Software Engineering

Contact

• Blog
• GitHub