Sabitlenmiş Tweet
发一下中文版的demo。
Orbital是一个永不失忆的agent - 自带一个本地的project。agent会自己更新，维护和参考project里的知识，保证每一次新会话都是基于最新的工作进展，不需要每次都给agent交代任务背景，解释预期和要求
youtube.com/watch?v=-Nn0Dk…
YouTube
中文
Keane
301 posts
Keane
@keane42443
Architect in AI. PM at Tencent. Orbital - the agent that never starts from zero https://t.co/wHsdmUbH4h
Katılım Mart 2026
78 Takip Edilen31 Takipçiler
@arkuy99 说实话不如用cowork让文件直接放本地，或者直接用orbital，用本地文件夹作为project和沙箱，然后claude codex，插个llm apikey的其他agent都能用。
github.com/zqiren/Orbital
中文
when coding is fully solved, many early-career software engineers will be forced to completely pivot their careers and become forward-deployed software engineers or solutions architects or product marketing managers or AI engineers or quantitative data scientists
English
pi 也有自己的computer-use 了， 支持Mac 和Windows。
看起来是作者逆向了codex 的实现，底层不是cua driver。
我觉得稍微改造一下，应该就可以给任意harness 使用。
github.com/injaneity/pi-c…
Zane Chee@injaneity
pi-computer-use v0.4.3 is here! an update to bring us to parity with codex (who to their credit, stepped up their game massively in their last update): - a ghost cursor now shows what computer use is working on - parallel tasks allow subagents to coordinate across different apps safely - better batching improves the accuracy of repeated / simple groups of actions we now see an almost 50% improvement in latency across a variety of tasks, and better handling for non-trivial use cases (like microsoft apps) try it out here: github.com/injaneity/pi-c…
中文
@alex_prompter bro that was actually the perfect model. no new model till today give me the feeling of the initial opus 4.6. the spike from 4.5 to 4.6 is crazy
English
@QuantumTransf 是的！啪一下一排agent。但我个人喜欢用sub agents。要是task spec比较长，或者需要和主agent对齐design的话（一般很多轮），context rot感知特别明显，得用subagent 来整
中文
然后今天就看到很多 superpowers 受害者，一个 ultra 生一堆 subagent（
你的 agent 如果有这种症状的话，可能需要看看电脑上有没有 subagent-driven-development / test-driven-development skill
检测到您可能是 superpowers 的受害者！
Yuu💖@QuantumTransf
体验下来似乎 gpt-5.6-sol 更喜欢用 skill 了？（就是根据 meta 头去匹配用什么 skill） 感觉是好事，给一些有 cli 的 skill 和 tool 同等的地位了，不过对于 skill 装很多的人来说可能不那么好（
中文
here here!
orbital is a local agent that lives in your local folder. it maintains a "project" in the local folder in which all transcript, distilled skills, artifacts and long-term memory is stored. it also supports dispatching claude code and codex as sub agents.
github.com/zqiren/Orbital
English
i'm so sick of using claude code in a terminal
i'm not coding. who has made a great app that i can use with any model?
English
真的很离谱，所以我觉着harness还是开源的好，并且一个重要功能一定是沙箱，得把控制权还给用户。
拿既然说到了这里！推荐一下自己做的open source harness🤡🤡：github.com/zqiren/Orbital
开源的，所有agent运行的结果和transcript，以及长期记忆全部存在本地文件夹里 - 以“项目”（也就是本地文件夹）边界作为沙箱来管控agent行为，支持claude code codex
shameless self promo lol🤡🤡
中文
this removing 5 hr limits is actually crazy. i always though having the 5 hr limits is to even out usage so as to control the amount of parallel api calls at the same time. in this way they can reduce the inference cost of machine by preparing for a reduced max.
but openai just said fk it, it doesnt care...
English
Anthropic: "We'll give you peasants Fable for another week at 50%"
OpenAI: "We made a tiny error. We're going to remove the 5hr limit so you can go nuts. OH! We're also going to reset again for the 4th time in 48hrs."
English
@dexhorthy @vaibcode wow. im actually not sure if i should be scared or amazed..........
English
@keane42443 @vaibcode what’s crazy is that was a random image gen from an ai tinkerers Paris - nobody put me in the prompt im just in the weights
English
if you wanna really cut the slop from AI code, you gotta focus on not just architecture but **program design**
If you think architecture alignment is enough to get slop-free code, i got news for ya - a couple mermaid diagrams ain’t gonna cut it
shouts out to @vaibcode for leading another dope epsiode
English
我觉得很多人在这里想多了。
在我看来，Anthropic 根本没那么担心消费者流失。
原因很简单：
C 端订阅可能看起来热闹，但真正赚钱的不是这里。
Claude 的个人订阅其实一直是补贴状态。
真正的大头在：
企业客户
商业 API
大规模 Agent 部署
这些客户愿意为稳定、能力和安全性支付高得多的价格。
这也是 Anthropic 真正领先 OpenAI 的地方。
所以不要觉得 Dario 会因为一些用户取消 Max 订阅就睡不着觉。
Fable 5 限制个人访问，更像是在重新分配算力：
把最贵的计算资源留给付费能力更强的企业场景。
对 Anthropic 来说：
少一点高成本 C 端用户，可能反而是一件好事。
真正的战争，不是在谁的订阅用户更多。
而是谁能拿下企业 AI 的长期入口。
Ali Haider@ggg78g89
🚨SCOOP: MY Friend at Anthropic says things are VERY tense internally. Dario's running tough meetings — GPT-5.6 Sol is strong and Grok 4.5 is right on Opus's heels. Pulling Fable from subs on July 12 would trigger mass cancellations (why keep Max for Opus 4.8?), so they're now pushing to keep Fable 5 in subs permanently.
中文
现在各大 AI 厂商疯狂抢用户，使用各种重置、加大限额的这种百亿补贴的方式。但有没有想过，如果有一天百亿补贴不再有了，那我们真的还用得起吗？
Tibo@thsottiaux
Morning. The last 48 hours of Codex and ChatGPT Work have been intense! Three important updates: - Temporarily removing the 5 hour usage limit restriction for all Plus, Business and Pro plans - Rolling out changes that will make GPT 5.6 Sol more efficient across the board and that will be reflected in less usage being used so that it can take you further. Exact impact to be quantified and shared - We hit 6M active users, and are landing a usage reset in the next hour Go do things
中文
