Ask HN：你们用什么规范来驱动 AI 智能体？
3 分•作者: midnight_eclair•5 个月前
我目前的信念是：
1. “自主”开发已成定局，无论我们是否喜欢。
2. Markdown 提示是驱动此类开发的糟糕方式，无论你试图施加多少结构和层级（无论是链式还是分叉到不同角色等）。
3. 拼图式方法没那么糟糕——用你选择的规范性语言声明任何组件或子系统的“边界”，然后让智能体填充内部。
如果你认为这些信念是错误的，没关系，但我更感兴趣的是那些大致同意的人的反馈，关于最后一点：你用什么规范性语言来描述组件边界、功能和跨组件交互？
我用 OpenAPI 取得了一些微弱的成功，但也许有更好的替代方案？
查看原文
my current convictions are:<p>1. "agentic" development is here to stay, whether we like it or not<p>2. markdown prompts are a terrible way of driving such development, no matter how much structure and hierarchy you're trying to impose (whether chaining or fan-out into personas, etc)<p>3. jigsaw-puzzle approach is less terrible - declare the "boundary" of any component or subsystem in a normative language of your choice and let the agent fill the insides<p>it's fine if you think those convictions are incorrect, but i'm more interested in feedback from folks that more-or-less agree, with regard to the last point: what do YOU use for normative language to describe component boundary, function and cross-component interactions?<p>i've had mild success with openapi, but maybe there are better alternatives?