Using Moli and Codex to turn Quora threads into research leads
In a hands-on report, 少数派 AI showed Codex using Moli Web Fetch Skill to retrieve material from a Quora “Technological Innovation” topic and organize it into a research list. After an initial Markdown extraction returned only an error, an alternate network-response path retrieved 10 answer cards, one of which provided a possible 3D-printing angle. The author said the cited faucet-washer account still requires verification.
The report demonstrates a concrete AI-agent research workflow in which Codex orchestrates Moli, recovers from a failed page extraction, and turns retrieved web content into leads without manually copying the material.
undefined
“技术创新”这个题太大了,直接写很容易空。我想找点具体材料,就把问答社区 Quora 上“Technological Innovation”这个话题的链接交给了 Moli。结果,在取回的 10 条回答里,一个水龙头垫片让我找到了想往下写的方向。
我用的是 Codex + Moli:我给链接和要求,Codex 安排步骤,调用 Moli 去网页里取内容,再把资料整理给我。Moli 是一款给 AI 用的开源浏览器,可以在本地运行。这次我想试的,就是把“读网页、取材料”这段工作也交出去,不用自己先复制一堆内容,再让 AI 总结。

Moli官方产品图|面向AI Agent的开源浏览器。来源:lexmount/moli官方仓库。
贴一个链接,让 Moli 把网页读回来
读网页这一步,我用的是 Moli Web Fetch Skill。它是一套给 Codex 的操作说明,告诉它怎么调用 Moli 读取页面、等待内容加载、检查结果。装好 Moli 和这个 Skill 后,我在 Codex 里选中 Moli Web Fetch,贴上话题链接,只加了一句“View all content”(查看全部内容),没复制回答原文,也没预设表格。