Maggie Appleton 写设计、人类学和智能体界面时，习惯先画清楚隐喻，再谈工具。把「How to Import Academic Papers from Zotero into Tana」整理成可阅读的中文笔记：问题在哪、界面默认会把人带去哪、落地时该改什么。原站导航、评论回响和广告已去掉。
How to Set it Up
I use Tana as my primary knowledge base. It’s an outliner with a powerful structured data system. Each node in your workspace can have a distinct type like “Person,” Book,” or “Article” and pre-set properties on those types like “Author,” “Abstract,” or “Reading Status.”
An example daily notes page in Tana with typed nodes The Tana team developed a special format for getting structured data into Tana from external sources (your browser, Twitter, Readwise, etc.). It’s called Tana Paste . They have a bunch of platform-specific examples here .
值得单独记下的观察
- Download my translator file from this GitHub Gist . This is a JavaScript file that converts the Zotero metadata into Tana Paste format.
- Select a paper in your Zotero collection and hit Cmd+Shift+C to copy to your clipboard it in Tana format. Alternatively, right-click and select Export Item. You’ll need to select “Tana Metadata Export” again from the dro
- Open Tana, make a new node, and paste the item in. It should look something like this:
落地时建议先做的 5 件事
- 先写清这个工具在强化思考还是在替人思考。
- 默认交互不要只会谄媚：该追问、该给反例、该要求证据。
- 智能体界面要露出推理步骤，而不是只给一个光滑答案。
- 知识发布用可生长的笔记，而不是一次性营销长文。
- 改产品前先画隐喻：用户以为自己在做什么，系统实际在做什么。
和智能体产品怎么接
龙虾PRO做 OpenClaw 落地时，最该从这类笔记里拿走的是「别把聊天框当唯一界面」。数字员工要能追问、能验证、能把过程摊开，而不是只负责说好话。
本文侧重智能体生命周期与技能边界。若你要动手验证，优先用 Playground 做小任务压测；权限与托管再单独规划，避免「先装一堆再治理」。 相关：Playground · 创建智能体 · 学习中心
常见问题 FAQ
什么是知识与智能体怎么？
「知识与智能体怎么」可概括为：I use Tana as my primary knowledge base. It’s an outliner with a powerful structured data system. Each node in your workspace can have a distinct type like “Person,” Book,” or “Art 本文从定义、方法与实践要点展开说明。
为什么要关注知识与智能体怎么？
关注知识与智能体怎么，是因为它直接影响效率、风险与可复制性。文中指出：I use Tana as my primary knowledge base. It’s an outliner with a powerful structured data system. Each node in your workspace can have a distinct type like “Person,” Book,” or “Article” and pre-set properties on those types like “Author,” “Abst…
如何落地知识与智能体怎么？有哪些关键步骤？
建议按以下路径推进知识与智能体怎么：1) Download my translator file from this GitHub Gist . This is a JavaScript file t…；2) Open Tana, make a new node, and paste the item in. It should look something lik…；3) 先写清这个工具在强化思考还是在替人思考。；4) 默认交互不要只会谄媚：该追问、该给反例、该要求证据。；5) 智能体界面要露出推理步骤，而不是只给一个光滑答案。。细节见正文对应章节。
知识与智能体怎么适合哪些人或团队？
知识与智能体怎么更适合：产品/技术负责人、运营与增长团队、需要落地智能体或自动化的中小团队、关注「AI智能系统」方向的读者。若你只需要单次聊天式问答，可先读概念；若要上生产，请重点看步骤、权限与风控相关段落。
关于「How to Set it Up」，本文给出了什么结论？
在「How to Set it Up」部分，要点是：ook,” or “Article” and pre-set properties on those types like “Author,” “Abstract,” or “Reading Status.” An example daily notes page in Tana with typed nodes The Tana team developed a special format for getting structure
关于「值得单独记下的观察」，本文给出了什么结论？
在「值得单独记下的观察」部分，要点是：t+C to copy to your clipboard it in Tana format. Alternatively, right-click and select Export Item. You’ll need to select “Tana Metadata Export” again from the dro Open Tana, make a new node, and paste the item in. It sh