stringer fan

@0xSero I am thinking if this feature can be summarized to a skill that I can use it in Claude code or codex

This is the best feature Factory has made so far, I have been prompting models to build me visual-explainers for a long time now To learn a subject/repo/research I put the files in a repo and run autowiki to keep it updated. Of course you have to be careful models hallucinate

@FactoryAI lovethis

Introducing AutoWiki in Factory.

@Vlnrcm @gintokijitui @MercedesAMGF1 @GeorgeRussell63 Go back to your desert🤣🤣🤣

Writing his name into Mercedes racing history 🩵 George joins the 💯 club. He’s now only one start behind VB in the all-time list 👊

@MercedesAMGF1 @GeorgeRussell63 令人恶心的职场霸凌，梅赛德斯真的是糟糕透了。

@ninthbit_ai @jaywcjlove 我一直以为是我代码（怀疑是宽度冲突）的问题。。。 我写的软件怎么都解决不要这个问题，然后我干脆把软件窗口的最小值写死在一个较大的值了。

@jaywcjlove 之前这个切换按钮还让我遇到了性能问题，在宽度小于一定值时，侧边栏从隐藏状态变为展开状态，动画会非常卡。但是当我宽度拉大时，就一点也不卡了。至今不知道为啥，我把右边屏幕中的列表改成懒加载之后会好一点。

侧边栏样式回到了 macOS 15 幸幸苦苦的，隐藏侧边栏“切换按钮”， 在 macOS 27 默认给你去掉了 😁

@DevBySami What’s this

Lots of people been asking for an agent-native tmux. Working on this. Drop some ideas!

@sqs So, at least based on Amp’s own published numbers, the claim that the expensive frontier model is “usually the fastest and cheapest” does not hold without very specific qualifications around task type, reasoning level, pricing assumptions, and evaluation methodology.

@sqs Your statement — “On these real tasks, the expensive frontier model was not only the best, but also usually the fastest and cheapest” — seems to directly conflict with Amp’s own GPT-5.5 model page.

The Factory team seems very smart, and I'm eager to see how well their cost-optimizing new model router works. If they can pull it off, that is a huge accomplishment. For Amp: we may try to build a model router that picks the best model for a task, but we don't intend to build a cost-optimizing model router based on the current state of the models. Here's why. Every time we've looked into using cheaper models in Amp, we've benchmarked on tasks that reflect how people use agents for coding today. On these real tasks, the expensive frontier model was not only the best (obviously), but also usually the fastest and cheapest, when measuring end-to-end task completion. Why? Cheaper-per-token models are less capable, which means that on complex real-world tasks they spend more tokens and time fixing mistakes along the way. You can find plenty of cases where cheaper models are indeed faster and cheaper end-to-end. But such cases were rarer than we expected, and the differences were fairly small. If you can easily detect such cases, then there is an opportunity here. But even then, on the AI hedonic treadmill, once people get a taste of frontier intelligence, they don't want to go back to using those more primitive prompts where cheaper models suffice. (Which is a good part of human behavior! It's how we decided to stop living in caves!) If your tasks can be handled just as well by non-frontier models, I would strongly advise you to uplevel how you use agents and what you produce to stay competitive against people who are using frontier models. In a power-law world, with rapid intelligence advances, try to get to the frontier and stay there.

@daniel_mac8 ampcode has it loooong time ago

Killer new Codex feature that went unannounced: Codex can now coordinate threads for local projects/worktrees. Includes separate background threads. Transforms Codex into a meta-agent that can orchestrate its own workspace.

@thsottiaux That’s very nice, love codex more than Claude code

A little secret. About 5% of our production traffic is on the Pi harness, about another 5% is on OpenCode. Reminder you can use your ChatGPT account in a flourishing set of other tools. We’ll continue to make Codex awesome, but you have options.

@sama frontend plz

what would you most like to see improve in our next model?

@sqs @united Hi! I’m flying tomorrow morning and would love to help test. Amp username: aurelian Route: Singapore, then Malaysia over the next week, and back within 10 days. Happy to use Amp during the flights, post feedback from in-flight WiFi

The final boss of an interactive client<->server application is intermittent airplane wifi. Amp worked pretty smoothly for me on a 15-hour SFO-SYD @united flight, but we want to make it buttery smooth. If you have a flight in the next 10 days and want to help us test, reply here with your Amp username and flight route, and I'll send $250 in free credits. You then just need to post your feedback using Amp in flight. Would also love a photo of Amp in flight, and a comparison to how other agents work in flight. (I'll have Amp decide for me which of the replies seem legit. I reserve the right to not pick everyone.)

@xin_pai88825 期待一波

@Aurelian_Ari 有计划，我看下怎么处理比较好，因为我当时做了非常多的功能，但是太乱了整个项目

去年 9 月份跟Para 翻译的开发聊关于提示词变量工程（图一），虽然我设计稿都做好了，可惜他太忙了没时间给我处理。 在年底的时候，我把这个功能做到了我的桌面 agent 上（图二），之前跟多层级的布尔绑定。但因为做到一半实在没精力做下去了，我放弃了那个项目，最近我又把这套理念放到了一个生图项目里，发现无比契合。用户不用写复杂的提示词，只要通过我预设好的提示词（图三），把配置项选一下，即可生成高质量的图片

@sqs I'm new to Ampcode and I like the idea handoff is better than compact. , but Ampcode doesn't support auto-handoff, my codebase is big, some times only one task can nearly run outoff context window(in deep mode which I perfer) before manually handoff, how handle this problem?

I've had a very good experience with Opus 4.7 all along in Amp's `smart` mode. If you disliked Opus 4.7 in other agents, try it in Amp.

@thsottiaux Codex is better than Claudecode

I don't know what they are doing over there, but Codex will continue to be available both in the FREE and PLUS ($20) plans. We have the compute and efficient models to support it. For important changes, we will engage with the community well ahead of making them. Transparency and trust are two principles we will not break, even if it means momentarily earning less. A reminder that you vote with your subscription for the values you want to see in this world.

【开工季福利来袭！！！】 为了感谢大家对我们的支持，我们打算免费送一台 Mac Mini256GB + 2000刀PackyAPI额度，助力2026年VibeCoding 和养殖龙虾！ 一等奖：macmini 256GB 一台 二等奖：Packyapi额度100 刀✖️10 人 三等奖：Packyapi额度50 刀✖️20 人 点赞+评论+转发即可参与抽奖！(中奖后我们会仔细检查哦) 3 月 8 号截止～ 网站链接：packyapi.com 相关文档：docs.packyapi.com

@Murphychen888 M大，这个投资者信心指数是咋计算的？😈我想放到我的策略里，这个东西太管用了

我从未见过“投资者信心指数”在零轴以上的犹豫区（黄色）停留这么长的时间...... 在过往的数据中，短则1周长则1个月，一定会出现新的方向，要不向上进入绿色的乐观区，要不向下进入红色的恐慌区。而这一次，从8月27日至今已整整49天！在历史上也绝无仅有。 我们知道这是个大级别的趋势指标，信号并不敏感，但有极高的确定性。每2段红色区中间的绿色区，对应的就是一轮大级别趋势行情，所以一旦进入红色时段，就是趋势交易者的最佳击球区（可以重仓）。 （图1） 但是现在指标始终停留在黄色段，代表当前市场投资者的情绪正在不断摇摆，对于方向并没有产生一致性的共识。因此，在这个位置，我们很难赚到有确定性的大趋势的钱。 当我看不清时，首先就会控制仓位，留足现金，耐心等待最舒服最明确的位置出现。至少现在还没有进入红色段，说明从4月启动的这轮趋势还未结束，我们现在正身处其中一个上不上，下不下的位置。 那我为什么不空仓？因为牛市的基础还在。至少我个人观点并不认为现在就是大周期上牛转熊的开始（逻辑可以见以下推文链接）。因此，在有性价比的支撑位附近，我也会买入部分仓位的现货，可以在一定程度上避免踏空的焦虑。 x.com/Murphychen888/… x.com/Murphychen888/… 赚不到趋势的钱，就只能在震荡中找波段机会，或小级别短线机会。这就对交易者的综合素养提出了很高的要求。也正因此，才让很多小伙伴都觉得现在交易难度极大，钱难赚。BTC是如此，更别说山寨币了...... 先确保如何活的更久，再考虑如何赚的更多，这比什么都重要！ 以上仅用于学习交流，不作为投资建议 ------------------------------------------------- 本文由 #Bitget ｜ @Bitget_zh 赞助

Restoration of access to Z-Library and the cessation of illegal criminal prosecution - Sign the Petition! chng.it/52Yr5hch 来自 @Change