Why Read It Later Tools Fail
摘要：稍后阅读工具往往变成只写不读的归档库，队列不断堆积并最终被放弃。问题的根源不在意志力，而在输出端摩擦过高。解决方案是降低输出摩擦：通过摘要帮助判断是否阅读，翻译消除语言障碍，以及搜索让旧剪藏随时可被检索。本文同时是 Tiro 流水线的端到端测试，涵盖剪藏、摘要、标签、翻译、图片本地化和部署。
Why Read It Later Tools Fail
为什么稍后阅读工具会失败
Most read-it-later tools become write-only archives: articles go in, but nothing ever comes back out. The queue grows, guilt accumulates, and eventually the tool is abandoned.
大多数稍后阅读工具最终变成了只写存档：文章存进去了，却再也没拿出来过。队列越来越长，内疚感不断累积，最终工具被彻底抛弃。
The fix is not better willpower but lower friction on the way out: summaries that let you decide whether to read, translations that remove the language tax, and search that makes old clips discoverable exactly when they become relevant.
解决办法不是靠更强的意志力，而是减少取出时的阻力：用摘要帮你决定是否阅读，用翻译消除语言负担，用搜索让过往剪藏在恰好需要时能被找到。
const value = (article: Clip) => summary(article) + translation(article) + search(article);
const value = (article: Clip) => summary(article) + translation(article) + search(article);
This article exists to test the Tiro pipeline end to end: clip, summarize, tag, translate, localize images, deploy.
这篇文章的存在是为了端到端地测试 Tiro 流程：剪藏、摘要、打标签、翻译、本地化图片、部署。