展示HN：Tiptap AI工具包 – 现在探索服务器端自主文档编辑
我们最近发布了 Tiptap AI 工具包（<a href="https://tiptap.dev/docs/content-ai/capabilities/ai-toolkit/overview">https://tiptap.dev/docs/content-ai/capabilities/ai-toolkit/overview</a>），该工具包允许开发者将基于大型语言模型（LLM）的编辑功能嵌入到富文本编辑器中。
现在，我们正在探索下一步：让智能体在没有浏览器参与的情况下操作文档。
这个想法是建立一个服务器端系统，让智能体能够实时读取、写入和转换文档——本质上将文档层视为一个可编程的数据存储，智能体可以直接与之交互。我们内部开玩笑称之为“实时智能文档数据库”，因为它正朝着这个方向发展。
我很想知道其他人对这个方向的看法。
如果你有一个数据库，其基本单元不是行/对象，而是具有语义结构的文档，智能体可以对其进行修改，那么它还可以或应该做些什么？
有哪些明显的风险？
你希望在智能体的自主性和文档的完整性之间设定哪些边界？
我很想听听你们如何扩展或限制这样的系统。
查看原文
We recently shipped the Tiptap AI Toolkit (<a href="https://tiptap.dev/docs/content-ai/capabilities/ai-toolkit/overview">https://tiptap.dev/docs/content-ai/capabilities/ai-toolkit/o...</a>), which lets developers embed LLM-powered editing into rich text editors.<p>Now we’re exploring the next step: making agents operate on documents without a browser involved.<p>The idea is a server-side system where agents can read, write, and transform documents in real time — essentially treating your document layer as a programmable data store that agents can interact with directly. Internally we’ve been joking about the name “real-time agentic document database”, because that’s the shape it’s starting to take.<p>I’m curious how others would think about this direction.<p>If you had a database where the fundamental unit wasn’t rows/objects, but documents with semantic structure that agents can mutate, what else could or should it do?<p>Where are the obvious footguns?<p>What boundaries would you want between agent autonomy and document integrity?<p>Would love to hear how you’d expand or constrain a system like this.