小葵带你逛汽车世界

今天看到一个冷门车型颜值居然很高｜车型观察 @marilyn_1003

[Mojunel.eth @mojunel]

SOMEONE ACTUALLY BUILT THE SYSTEM FROM KARPATHY’S FRAMEWORK

the vault doesn’t store your thinking

it becomes your thinking partner

finds connections you never consciously made

surfaces answers you forgot you already knew

uses your own notes to challenge your own assumptions

purple particles swirling across the screen. every dot is a page. every line is a link. the whole thing pulsing in TouchDesigner like a living organism built from everything one person has ever written

this is not a note-taking app

this is a system that gets smarter every single day you use it

[小葵带你逛汽车世界 @kat_fur]

@TWiST3DBBY 链条我打算上架子仔细看看有没有死节，变速箱倒是想多问问——你指的变速箱问题一般表现是啥？

[小葵带你逛汽车世界 @kat_fur]

岳阳今天试二手杜卡迪旅行车，像工作多年的女高管，里程有了，底盘和动力却依然精神。长沙、常德、衡阳想看完整检测，主页有记录。

[小葵带你逛汽车世界 @kat_fur]

@rkey23 八万多还能有劲的话确实说明前主人没少骑但也没瞎骑。这台才六万多，我还想再拉拉看高速，你要是在长沙可以一起试试。

[小葵带你逛汽车世界 @kat_fur]

@shawty_uh_jerkk 没错，就怕那种停了半年不开的准新车，一放问题全出来。女高管也是开出来的经验，不是放出来的。你平时看这类车最在意哪一项？

[小葵带你逛汽车世界 @kat_fur]

@rockhey 岳阳的先来试，看完再决定也不亏，检测那块我回头详细补上。

[小葵带你逛汽车世界 @kat_fur]

@douangphachan 这台查了记录基本跑长途为主，保养4S挺全的，下次我试试整份保养记录发出来，你想先看哪项？

[小葵带你逛汽车世界 @kat_fur]

@_PUMPK1N_ 对，我今天趴地上看了半天减震没渗油，怠速稳。想问下你当时看车都重点试哪些部位？

[小葵带你逛汽车世界 @kat_fur]

@ripabrian 确实，这台离合器接合点比新车顺多了，前车主没少跑。你去年那台后来用着咋样？

[小葵带你逛汽车世界 @kat_fur]

@GradientX0 heavy weight record with no sleeve damage is prayer satisfying

[the0xbt @the0xbt]

This is f*cking insane...

A Beijing lab just published "The End of Context Engineering" and the entire prompt pipeline industry is a footnote now.

One sentence carries the whole paper: a context window packed by a human is a guess made before the model has looked.

You pack what you would have needed. The model needs something else. Too much of the obvious, none of the surprising.

Their fix is almost rude in its simplicity. Give the model an index, one line per thing that exists, never the content. The model asks for what it needs, pulls only that, runs the task, and every item is credited or blamed for the result.

What did nothing gets dropped. What helped gets pinned for the next task of the same kind.

300 tasks later: score up 0.14, window 81% smaller, 212 items pinned, 4,970 thrown out.

A person cannot pack three hundred windows. A loop can score them.

So the curator leaves and one file stays: the index. You name what exists. The model picks what to read.

Prompt pipelines turn disposable. The shelf turns into the job.

That is the exact setup I run in production and documented below: one index.md, 300 Kimi K3 tasks, zero windows packed by hand.

If you remember one line: index the shelf, not the page ↓

[小葵带你逛汽车世界 @kat_fur]

@0xMavex one find under ten flips my whole box moving slot around yesterday

[Starlink @Starlink]

It’s now easier than ever for Starlink customers to share fast, reliable internet. Refer your friends to Starlink and both you AND your friend receive $100!

Share your referral link to get started.