Umaru

[Umaru @kira8565]

@klarkxy zcode 还可以，codex 经常会过度防御

[Umaru @kira8565]

双机 dgx spark glm flash 5.3 155decode... 震惊瘫坐，核弹爆炸
#dgx #glm #tensorfold

[Umaru @kira8565]

@LinearUncle 这个做的很棒

[Umaru @kira8565]

@manateelazycat 用 dgx spark 也测了一把，没见有多大提升，dgx 小伙伴可以避坑了

[Umaru @kira8565]

GPT and Grok have zero design taste, while Gemini has great aesthetic sense but completely falls apart when turning it into actual code.

Here’s the cheat code: open Gemini in Chrome, have it look at the page and draft a detailed UI prompt, then feed that prompt straight into Grok. Works like magic.
✨✨✨

#gemini #grok #chatgpt

[Umaru @kira8565]

gemini flash 能力很强，惊呆了

[Umaru @kira8565]

@Ion_Mio_ @Seefs_ prefill 和 decode 能到多少呀

[Umaru @kira8565]

@realchendahuang Gemini Flash 在UI的品味上还可以，但是要去Antigravity里面用，接到Codex效果不是很好

[Umaru @kira8565]

@ai_xiaomu 亲测可以用Grok Bot分摊一些精力，体验不错的。让他去管Cursor😀😀

[Umaru @kira8565]

Cursor Ultra的Grok档位，非常的好用，量大，快，智商在线，非常不错

[Umaru @kira8565]

Claude: 不欢迎你们，给我爬[发抖][发抖]

GPT： 欢迎欢迎，但是我给你的 token兑水的时候你得接受，我这也忙，token 出的慢[旺柴]

Grok： 我比他们都贵，没钱的别来，模型追上来了，应用层的课还没补上，但是不折腾。 另外出（正经）图的体验会差点

Gemini： 有冤大头来了呀[旺柴][旺柴]

[Umaru @kira8565]

@Finding8964 @grok 事实核查

[Umaru @kira8565]

@elonmusk grok build desktop with remote control like codex

[Umaru @kira8565]

all in grok !

[Umaru @kira8565]

梁叔叔的DeepSeek API没有自媒体们讲的那么便宜，7分钱一个游戏啥的。。开了一个不是特别长的任务，一个6个stage，已经干到4块去了，不算返工，光这个任务干完估计就能到8～10块

[Umaru @kira8565]

cursor 的canvas 体验一流，离不开了

[Avi Chawla @_avichawla]

Prompt, context, harness & loop engineering, clearly explained!

An agent is a while loop with four layers of engineering wrapped around it:

- Prompt engineering
- Context engineering
- Harness engineering
- Loop engineering

Each one wraps the last, and the model sits in the middle, so none of them compete with the others. Instead, they just zoom one level further out.

> Prompt engineering:

This defines the input the model sees on one call, often composed of a role, instructions, examples, and an output format.

The techniques here alter the internal computation and reasoning the model goes through due to the wording it sees:

- Chain-of-thought makes it work in steps before answering
- Few-shot examples define the format and the edge cases
- A JSON schema or XML tags make the output parseable by code
- Self-consistency samples a few chains and takes the majority

> Context engineering:

It's everything the model sees on a turn, not just the prompt. That includes the query, retrieved docs, memory, prior turns, and tool outputs from earlier steps.

The window is finite and fills up fast, so the engineering work is to rank inputs and cut everything that isn't pulling weight.

You do this by:

- Retrieving only the chunks relevant to the query, then reranking them
- Keeping key facts out of the middle, where accuracy drops
- Summarizing old turns, evict stale outputs, push big blobs to files

> Harness engineering:

It's the code around the model that defines the tools, parses the calls, retries on failure, and can route work to sub-agents so one handles retrieval and another handles code.

A verifier then grades the result by running tests, validating a schema, etc.

Prompt and context involve getting one call right. The harness involves everything that has to happen around that call for it to run in a real system.

> Loop engineering:

In the usual setup, you manage the outer loop, i.e, you write a prompt, read the turns the agent runs, write the next prompt, and repeat, while catching failures.

This layer hands that job to the agent itself. It kicks off on a schedule or an event, and runs many turns with no prompt in between.

A loop inherently doesn't know when it's finished. An agent can report that it's done and halt while the tests still fail. So the stop can't be the agent's word, but rather it has to be a real signal, like:

- A turn and token cap to stop stuck runs
- A no-progress detector to catch repeated calls
- A completion check to verify the goal with a separate model or a deterministic test

By this layer, you're operating on the whole run, so the engineering moves from writing each prompt to setting the goal and the stop conditions up front and letting it run.

If you want to dive deeper into loop engineering, my co-founder wrote a full breakdown of that outer loop.

It goes from the basic while loop to a run that finishes on its own, with the code behind each part, and the parts that are hard to get right, like knowing when to stop, context rot over a long run, and keeping the checker separate from the maker.

Read it below.

[Umaru @kira8565]

一阶： 通过文档防腐，故事板驱动，测试护栏。让coding agent 具备稳定的长程任务能力

二阶： 通过构建 自主巡航，对抗循环，自主学习模式，让coding agent具备大型循环工程能力。让human in the loop 变成 human on the loop

太有趣了

[Umaru @kira8565]

@recatm 美式豆包

[Umaru @kira8565]

@viviennaBTC 图四懂王最爱香槟金，gpt出品