Coconut

[meng shao @shao__meng]

AI 写代码的速度上去了，但评审、审批、交接还是原来的节奏，效率提升在流程里被消耗掉了。

Anthropic 把六个阶段重新设计成一条产物链闭环：
每个阶段提交一个文件，下一个阶段读它开始；人只在门口审阅，Agent 做到生产门之前，不能越过去。 https://t.co/Z2An2o1GPC

[Coconut @watermelon_369]

@dhh You were made for this. 🙏

[AI超元域 @AISuperDomain]

AI 生成的架构图，终于可以不用再忍受“圆角方框 + Mermaid 味”了。

发现一个很有意思的开源项目：diagram-design。

它专门为 Claude Code、Codex、Pi 提供高质量的「编辑级」图表生成能力，覆盖架构图、流程图、时序图、状态机、ER 图、象限图、甘特图、数据流等二十多种类型。

最爽的是：

• 直接生成自包含 HTML + SVG
• 自动读取你的网站，匹配品牌配色和字体
• 支持 Light / Dark / Editorial 风格
• 可以把 Mermaid、https://t.co/VFffuVlDPB 直接重新设计
• 支持导出 SVG / PNG，用于博客、PPT、Figma
• 强调信息密度和视觉层级，而不是疯狂堆方框

如果你经常让 AI 帮你画技术架构图，这个项目非常值得装进 Codex / Claude Code。

GitHub：https://t.co/bVEsPWSomf

[Michael Dell 🇺🇸 @MichaelDell]

If you’re curious and love to learn, there has never been a better time to be alive.

[Coconut @watermelon_369]

@Suryanshti777 Roadmap

[Uncle Bob Martin @unclebobmartin]

Morning bathrobe rant: 20X.

[Matt Pocock @mattpocockuk]

100%

We are all engineering managers now

Strategists, not tacticians

[Kun Chen @kunchenguid]

most slop now comes from under-specification

it’s the human not really knowing what’s good, or failing to articulate it. AI then fills the blank with the most average assumptions

people who blame AI for average results should reflect on whether their input is truly novel

[Mario Nawfal @MarioNawfal]

🇳🇴 One decision Norway fans may never forget

New angle shows Alexander Sørloth choosing NOT to square it to Erling Haaland in a golden attacking moment.

That pass could’ve changed the entire game… and fans are losing their minds over what might have been.

Writer: Val

[Ryan Janssen @ryanjanssen]

There are only 3 models:
- Luna/High
- Terra/Ultra
- Sol/High

[周览资源 @grgerwcwetwet]

每次讲系统架构，最费时间的不是讲。

而是画图。

自己用 Visio、Draw. io一点点拖组件，不仅耗时间，还容易画得乱七八糟。

最近在 GitHub 上发现一个很有意思的项目 archify，把它装进 Claude Code、Codex CLI 或 opencode 后，直接用自然语言描述需求，就能生成架构图。

GitHub：https://t.co/f9AEvfSy4C

[Rory Flanagan @Rory_Flanagan]

The *actual* penalty specialist we left on the bench 🙃

[枫影 Justin Yan @MapleShadow]

博客更新: 为什么我用mattpocock/skills替代了superpowers

最近我用@mattpocockuk 的skills替代了之前在用的obra/superpowers，本文是我的一些想法。

https://t.co/IaMuvTZIaS

[Reads with Ravi @readswithravi]

I’m in love with this sentence:

“Pressure is a privilege. It means things are expected of you.”

[Santiago @svpino]

Context engineering is the single most important area you can focus on right now.

We already have amazing models.

Agents no longer fail because models are dumb. They fail because they don't have the right context.

Here are the 4 ingredients of good context:

[ThePersistence @ScottPresler]

We warned them — peacefully.

🧾

[Suryansh Tiwari @Suryanshti777]

Your Obsidian vault is probably dead.

Not empty.

Dead.

Because saving information is not the same thing as building intelligence.

Right now millions of people are spending hours building “second brains” that never think back.

Beautiful folders.
Perfect tags.
Endless highlights.

And absolutely zero cognitive leverage.

The scary part?

AI is about to make this gap brutally obvious.

Most people will use AI like a smarter Google:
ask question → get answer → forget answer.

But a small group is building something way more dangerous:

AI connected to years of personal context.

Their notes.
Their ideas.
Their unfinished thoughts.
Their reading history.
Their obsessions.
Their patterns.

At that point AI stops acting like a chatbot.

And starts acting like a cognitive extension of the person using it.

That changes everything.

Because the biggest advantage in the AI era will not be prompts.

It will be accumulated context.

People with connected knowledge systems are going to think faster, write better, learn quicker, and spot opportunities earlier than everyone else.

Not because they’re smarter.

Because they stopped starting from zero every day.

Most vaults are storage systems.

The best ones become intelligence systems.

[Kevin Rose @kevinrose]

insane...

[Bindu Reddy @bindureddy]

🚨 OPEN SOURCE AI IS LITERALLY UNSTOPPABLE 🚨

The legendary founder of Redis (Antirez) just dropped ds4 - a custom native inference engine built specifically for DeepSeek v4 Flash

This is earth shattering! Here is why:

DeepSeek v4 Flash is a quasi-frontier model with a massive 1M context window

You can now run it LOCALLY on a 128GB Mac using specialized 2-bit quantization

The architecture is reimagined—he moved the KV cache from RAM directly to the SSD disk! 🤯

We already know DeepSeek v4 Flash is insanely good for agentic loops - Now you don't even need the cloud to run it

Closed-source labs are burning tens of billions on massive GPU clusters while single brilliant developers are running frontier-level AI on laptops!

They told us open-source would be worthless against trillion-dollar monopolies

Instead, pure hacker culture + incredible open-weight models are completely rewriting the rules

Open Source will ALWAYS win 💕

[Thariq @trq212]

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.