I'm sad it took me this long to actually type /design but working on frontend with Claude Design and Claude Code is actually so good
我很遗憾自己花了这么久才真正敲下 /design，但用 Claude Design 和 Claude Code 做 frontend（前端）开发，体验真的非常好
TH
每条推文 hover 显示单独 ▶
this was a bug that was only live for a few minutes, ran into it on my personal account while doing some late night coding though
这是个只上线了几分钟的 bug，不过我昨晚深夜写代码时，还是在我的个人账户上碰到了它
if you've run into this, please restart Claude Code, fix should be propogating
如果你遇到了这个问题，请重启 Claude Code，修复应该正在逐步生效。
working more on a post about what we learned doing this and how you can apply this to your skills and system prompts
我们也在继续完善一篇文章，讲这次实践中我们学到了什么，以及你如何把这些经验应用到你的技能和 system prompts（系统提示词）中。
This was due to a heroic effort by many people at Anthropic working sometimes literally around the clock It was not at all clear that we'd be able to do this in time, and so proud of everyone who made it happen. Enjoy Fable.
这要归功于 Anthropic 许多人的一次堪称 heroic 的努力，他们有时真的是昼夜不停地工作。我们当时完全不能确定自己能否及时做到这一点，因此我为所有促成这件事发生的人感到无比自豪。请享受 Fable。
building prototypes of mockups, schemas, data models, proof of concepts, etc. is the best way to avoid spending tons of tokens before realizing you don't want the output
构建 mockups、schemas、data models、proof of concepts 等原型，是避免在意识到自己并不想要那个输出之前就花掉大量 token 的最佳方式
ideal prompting technique is: - thin prompts - thick artifacts + context - thin skills
理想的 prompting technique（提示技术）是：- thin prompts（精简提示）- thick artifacts（厚实产物）+ context（上下文）- thin skills（精简技能）
software engineering is the profession of automation
software engineering（软件工程）是自动化这门职业
this was such a fun talk with Cat and Simon, hope you get to check it out if you missed it
这次和 Cat 以及 Simon 的谈话特别有意思，如果你错过了，希望你能去看看
I'll open source this if it's interesting! But here's my first public artifact, a breakdown on my Mega Sceptile team:
如果这个东西有意思，我会把它 open source（开源）！不过这里先放出我的第一个公开成果：一份关于我 Mega Sceptile 队伍的详细拆解：
I've been playing a lot of Pokemon Champions recently and started using Claude Code to help me. It writes code using Smogon's npm library, pulls live usage stats and then writes reports to understand matchups, breakpoints or theorycraft teams.
我最近玩了很多 Pokemon Champions，也开始用 Claude Code 来帮我。它会基于 Smogon 的 npm library（库）编写代码，拉取实时使用率数据，然后生成报告，用来理解对局 matchup（对位）、breakpoint（关键阈值）或者进行队伍 theorycraft（理论构筑）。
This makes artifacts much more expressive and can be combined in creative ways. One of my favorites is to create a dashboard for a project in Claude Tag that can be edited by others or by my local Claude Code sessions.
这让 artifacts 的表达能力强得多，也可以用各种富有创意的方式组合起来使用。我最喜欢的一个做法，是在 Claude Tag 里为某个项目创建一个 dashboard，而且它既可以被其他人编辑，也可以被我本地的 Claude Code 会话编辑。
one of the core skills of agentic coding is reducing your unknowns
agentic coding 的核心技能之一，就是减少你的未知项
this has to be because coding agents change the engineering math on how it is to work with or port a legacy codebase, right? anyone at Riot able to confirm?
这肯定是因为 coding agents 改变了处理或移植 legacy codebase（遗留代码库）时的工程成本计算方式，对吧？Riot 有人能确认一下吗？
I'll be talking more about Claude Tag with @petergyang and at AIE with @_catwu. Let me know if there's anything you'd like us to dive into more!
我会和 @petergyang 一起进一步聊聊 Claude Tag，也会在 AIE 和 @_catwu 讨论它。如果有什么你希望我们更深入展开的内容，欢迎告诉我！
[1]
I'll be giving the the Day 2 opening keynote at AI Engineer World's Fair on July 1st at 9am. Hope to see you there, here's a quick preview
我将于 7 月 1 日上午 9 点在 AI Engineer World's Fair 发表 Day 2 的开场 keynote（主题演讲）。希望到时能在那里见到你，下面是一个简短预览。
Claude Code can now upload and edit HTML artifacts that you can share with your team or other Claudes! Starting with teams so you can share internally with your team, coming to Pro and MAX plans soon!
Claude Code 现在可以上传并编辑 HTML artifact（产物）了，你可以把它们分享给你的团队或其他 Claude！目前先从 teams 开始，这样你就可以先在团队内部共享；很快也会推出到 Pro 和 MAX plans！
Slack now renders HTML attachments instead of just showing it as text 😭🙏
Slack 现在会渲染 HTML 附件了，而不再只是把它当作纯文本显示 😭🙏
and the video for reference: (I didnt get to use the updated designs in time)
以及这个视频供你参考：（我没来得及及时用上更新后的设计）
here's the deck from this video if you want to go over it yourself: lmk if you have any questions!
如果你想自己再过一遍，这里是这个视频用的 deck：如果你有任何问题，lmk！
Lots of people asked how I used Fable to edit its own launch video so I made a video about that! TLDR it wrote a lot of code & tool calls to use transcription services, ffmpeg, do colorgrading, use the figma mcp, make remotion UI and render it. I didn't touch a video editor.
很多人问我，我是怎么用 Fable 来编辑它自己的 launch video，所以我专门做了一个视频讲这件事！TLDR：它写了很多代码并发起了不少 tool calls，去调用 transcription services、ffmpeg、做 color grading、使用 figma mcp、制作 remotion UI 并把它渲染出来。我完全没碰视频编辑器。
at Code w/ Claude Tokyo! say hi if you see me around
在 Code w/ Claude Tokyo！如果你在现场看到我，欢迎来打个招呼
Fable is a step-change in models, and I hope it changes how you work with Claude. More to come in a series of posts on how it’s reshaped our work, but the TLDR: it’s time to be more ambitious.
Fable 是模型能力上的一次重大跃迁，我希望它能改变你与 Claude 协作的方式。接下来我会写一系列帖子，讲讲它如何重塑了我们的工作；但 TLDR（太长不看）的结论是：现在该更有雄心了。
An app can be a home-cooked meal (2020) personal software was a bit early in 2020 but in 2026, it really can be as personal as a home cooked meal, or a handwritten letter
一个 app 可以像一顿家常饭一样（2020）；personal software（个人软件）在 2020 年还稍微早了点，但到了 2026 年，它确实可以像一顿家常饭或一封手写信那样个性化
How dynamic workflows allow Claude Code to handle whole new types of tasks
dynamic workflows（动态工作流）如何让 Claude Code 处理全新的任务类型
If this prompt feels well written to you, it's because Suzanne is a writer in her little spare time! You can read her short story, Mall of America here: It's one of my favorite short stories about the human condition that happens to involve AI.
如果你觉得这个 prompt 写得很好，那是因为 Suzanne 在自己一点点业余时间里也是位作者！你可以在这里读她的短篇小说《Mall of America》：这是我最喜欢的关于人类处境、同时恰好涉及 AI 的短篇小说之一。
Workflows are the biggest upgrade to Claude Code’s capabilities since skills and subagents. I dove deep into it with @sidbid to figure out best practices, examples and more. I’m particularly excited about the non-technical tasks it enables for Claude Code.
自从 skills 和 subagents 之后，workflows 是对 Claude Code 能力最大的一次升级。我和 @sidbid 深入研究了它，以梳理最佳实践、示例等内容。我尤其期待它为 Claude Code 带来的那些非技术类任务能力。
[1]
getting it access to data is important too, add in the gmail, google calendar, etc. connectors but I think people underestimate how much context they have in files
让它获得数据访问权限也同样重要，把 Gmail、Google Calendar 等 connector（连接器）也加进来，但我觉得人们低估了他们在文件里 actually 拥有的 context（上下文）有多少。
- image or video editing? write scripts - finances, tax work, etc? put in PDFs, write scripts, output HTML - medical advice? put in PDFs + data, output HTML - filling out paperwork? write scripts - creating a report? write HTML - making plans? write HTML
- 图片或视频编辑？写 scripts（脚本） - 财务、税务工作等？放入 PDF，写 scripts，输出 HTML - 医疗建议？放入 PDF + 数据，输出 HTML - 填写文书表格？写 scripts - 制作报告？写 HTML - 制定计划？写 HTML
the basic trick to using Claude Code for non-technical work is to put a bunch of files in a folder and tell it can write scripts + make HTML
使用 Claude Code 处理非技术性工作的基本窍门是：把一堆文件放进一个文件夹里，然后告诉它可以写 scripts（脚本）并生成 HTML。
continuing my HTML era, I had so much fun talking with Claire at Code w/ Claude about staying in the loop with long running agents
继续我的 HTML 时代，我在 Code w/ Claude 和 Claire 聊如何在长时间运行的 agent（代理）执行过程中持续掌握进展，聊得特别开心
okay this is going kinda viral and tbh my original text was kind of messy, so here's a second pass with the help of Claude: -- Implement <SPEC>. As you work maintain a running implementation-notes.html file that captures anything I should know about how the implementation diverges from or interprets the spec, including: - Design decisions: choices you made where the spec was ambiguous - Deviations: places where you intentionally departed from the spec, and why - Tradeoffs: alternatives you considered and why you picked what you did - Open questions: anything you'd want me to confirm or revise
好吧，这条内容有点要 viral（爆火）了，说实话我原来的文本也确实有点乱，所以这是在 Claude 帮助下整理出的第二版：——实现 <SPEC>。在工作过程中，持续维护一个 implementation-notes.html 文件，记录任何我需要了解的、关于实现如何偏离 spec（规范）或如何解释 spec 的内容，包括：- Design decisions（设计决策）：当 spec 存在歧义时你做出的选择 - Deviations（偏离项）：你有意不按 spec 来实现的地方，以及原因 - Tradeoffs（权衡）：你考虑过哪些替代方案，以及为什么选择现在的做法 - Open questions（待确认问题）：任何你希望我确认或修改的事项
as much as you spec there are always still ambiguities and unknown unknowns that come up and this gives the model a good out to make decisions but keep you in the loop
不管你把 spec 写得多详细，总还是会出现各种歧义和未知中的未知，而这样做能让 model（模型）在做决策时有一个很好的出口，同时又能让你始终 stay in the loop（掌握进展）
tired: I brought my work laptop and my personal laptop wired: I am dual wielding
累了：我带了工作 laptop 和我的个人 laptop，有线连着：我这是在双持呢
I've been using HTML for planning, speccing, exploration, code review, reports and a lot more.
我一直在用 HTML 来做规划、编写 spec（规格说明）、探索、code review（代码审查）、报告，以及更多事情。
Jarred tried rewriting Bun in Rust and it passes 99.8% of the existing test suite we're not being ambitious enough
Jarred 试着用 Rust 重写 Bun，而且它通过了现有测试套件的 99.8%，我们还不够有雄心。
You can also see example HTML documents I've generated here:
你也可以在这里查看我生成的一些示例 HTML 文档：
HTML is the new markdown. I've stopped writing markdown files for almost everything and switched to using Claude Code to generate HTML for me. This is why.
HTML 是新的 markdown。对于几乎所有内容，我已经不再编写 markdown 文件，转而使用 Claude Code 为我生成 HTML。这就是原因。
[1]
"everyday we're trying to obtain more compute to pass on to you, we're sorry if it takes sometime but we're going to acquire as much as we can" you heard the man
“每天我们都在努力获取更多 compute（算力）来提供给你们，很抱歉这可能需要一些时间，但我们会尽可能多地获取。” 你已经听到他本人这么说了。
my white whale is when CC sometimes looks like it hangs during big file writes, I think we found it 😭
我一直苦苦追查的问题是：CC 在写入大文件时有时看起来像是卡住了，我觉得我们找到原因了 😭
a big part of this is we're working on making the no flicker renderer excellent so we can ship it as the default renderer, more updates coming there but lots of other papercuts coming
这件事很大一部分原因在于，我们正在努力把 no flicker renderer（无闪烁渲染器）做到非常出色，这样就能把它作为默认 renderer（渲染器）发布；这方面还会有更多更新，不过也还有很多其他烦人的小问题在陆续处理
we're doing a lot more of this, hunting down some of the most annoying bugs in Claude Code let me know if you have any white whales
我们正在做更多这类工作，追查 Claude Code 里一些最烦人的 bug（缺陷）；如果你也有自己一直苦苦追查的问题，告诉我吧