Agent 工程的价值，不是把每一步都交给模型
The value of agent engineering is not giving every step to the model
A first-person argument about where agent engineers create durable value when general-purpose agents can already use tools and operate software.
Writing
我如何理解 AI 工程。
中文正文，英文摘要。每篇文章都从一个判断开始，再把它放回证据、项目和限制里。
打开文章 ↗
对 Agent 来说，Context 往往比 Workflow 更接近行为的根源
For agents, context is often closer to behavior than workflow
Why I treat context, memory, identity, and permissions as runtime foundations instead of assuming that orchestration alone explains agent behavior.
Workflow 与 Agent loop：不要用一个词掩盖控制权
Workflow vs agent loop: do not hide control ownership behind one word
A practical distinction between deterministic workflows, graph runtimes, and LLM-led control policies, based on comparing real agent architectures.
没有评估，Agent 系统就只有故事
Without evaluation, an agent system is only a story
Why evaluation, failure analysis, and evidence quality matter more than a polished demo when improving production-oriented agent systems.