服务器技术专家，曾就职于阿里云 ，有赞，外企，在线教育等； 👑 AI coding - 职业工程师 💻 只分享硬核AI编程工具，技巧等
中国
Joined March 2020
- Tweets9.8K
- Following1.5K
- Followers15.7K
- Likes34.4K
最近我对 Deeseek Harness（DSH）改观了，现在很推荐拿它做日常开发。主要有 3 个原因：
1. 模型接入省心了。通过 magpie，几乎不费力就能接入 OpenCode Go、Codex 等第三方模型服务，不用再单独买官方 API。
2. 产品成熟了。有了客户端，日常开发用下来，体验已经和 Codex 很接近。
3. 插件市场是杀手锏。连 bot 相关插件都有了，常用需求基本能覆盖；缺什么，也可以自己开发，或者拿现有插件改。
如果你之前试过 DSH，觉得还不够成熟，现在值得再试一次。
非广。
今天提名几个广泛使用、内容丰富、更新及时的 DSH 插件导航站吧～
dshfind.com/
dshmarket.com/
awesome-dsh-plugin.com/
其中 dshmarket 也可直接作为 DSH 插件安装到设置页中。
（不代表公司立场，不对非官方网站的内容负责。）
很多人都在自制 harness，但怎么判断自己做得怎么样？
一个简单的观察入口：Magpie → 用量 → 上下文。
图里第二个 anthropic，就是我自己的 harness，通过 Anthropic API 格式接入 Magpie。
我这个垃圾harness缓存命中率很低，必然有bug!系统提示词也偏长。
最近几乎每天都会刷一会儿红果短剧，突然冒出一个念头：
如果把姜萍的故事改成短剧，让她真的拥有数学天赋，一路顶着质疑，最后拿下阿里巴巴数学竞赛冠军。
中专女生、数学天才、逆袭夺冠、全网打脸——这不就是短剧里的顶配爽文剧本吗？
突然想到，梁圣的 DeepSeek，是不是根本没有运营部？
果然只要产品够硬，能和群众站在一起，就算派个金牌选手当公关群众照样喜欢（产品，不是金牌选手）。
反观有些公司，一会儿包装创始人通音乐，彰显公司“有文化”，套餐名字取得谁都不认识，让你自卑。
模型宣发又看不起国内KOL们，一定要找洋鬼子宣发，而且不实事求是，一定要说比opus5.5便宜更好用😂，你当群众都是傻子吗？
又有一名 Claude app 桌面版工程师离职，难道 anthropic 在大裁员吗？🤔
Yesterday was my last day at Anthropic.
I created Claude Code on desktop because I believe everyone can be a builder. Every feature request and bug report we got from this community was a gift that impacted millions of lives. I’m eternally grateful for those gifts, the product wouldn’t be what it is without you.
Anthropic is doing some of the most important work of our time. Contributing to that mission will forever be a career highlight for me. I never imagined I’d get to contribute to something so special in my lifetime.
Walking away from my best friends, a product I poured my heart into, and the mission was the hardest thing I’ve done in my career. But now I have some free time for GTA VI.
I'm already itching to get back to addressing all your feedback. I'll share what's next soon!
人类对magpie 的使用了解不足50%，magpie 隐藏了很多强大功能，例如可观测性(prompt数据采集）。
去下载langfuse代码，让AI 在本地把langfuse 跑起来，最后在magpie里配置参数（全程截图让AI指导你）。
于是在lanfuse里可以看到完整的请求记录，太强大了！方便调试！
github.com/langfuse/langfuse
claude code 的扛把子 thariq小哥正式确认自己离职了，但他自己没有说原因。
anthropic 这几天就要上市，在上市前跑路，按照我的个人职场经验，不得不猜测最大原因可能上市跟他没多大关系， 没有股份或期权，或者持有的很少，上市对他的实际收益有限，或者对利益分配失望透顶！
因为正常上市员工是不能马上套现的，如果猜错了，请指正。
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined.
I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top.
The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever.
My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply.
To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had.
For me, it is time for a bit of a break before diving into something new.
If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
海内外几乎所有人都在吐槽 kimi的套餐不耐用😮💨
脱离群众太久了，不知道群众的想法和世间其他厂商的玩法。
劲爆！Moonshot 疑似找了海外英语圈 AI 大 V 投放商单，为新模型 Kimi K3.1 Pro 造势。👀
That's how Kimi AI are promoting their new model Kimi K3.1 pro and I know all the influencers working with them and none of them are honest,I have posted about it myself but I didn't join this and never used their templates that they provide.
PROPAGANDA, I have more to share it's f**cked up I'm gonna have follow up for this
卧槽，kimi K3.1 pro 这么强吗。
opus5.5 就问你怕不怕，颤抖吧！
要不要冲 99 先占个位子？就是之前k3 的时候那个速度之慢让我现在还有阴影，比openAI gpt-6模型还慢。
感谢博主推荐，建议收藏这个开源的江山时序图，可以一眼看到每个朝代的国土面积和历史事件，没想到清朝国土面积这么大！
在线：deanchen.me/jiangshan-map/
开源地址：
github.com/DeanChensj/jiangs…
建议 Kimi 学学MiniMax, 原来MiniMax 的部分流量已经路由到美国的inferact公司的GPU 上， inferact 已经在帮MiniMax 扛实际流量了。
这样也解释了这次MiniMax 3.1 flash 模型能免费这么久，有这么多算力可以支撑。
Kimi请学学吧!给用户一个畅快模型推理体验。
Today, we are announcing a strategic partnership with @MiniMax_AI to advance frontier production inference for open-source models.
The Inferact Platform has served MiniMax M3.1 Flash in production since the day it launched. Requests through the MiniMax platform are routed to our stack alongside MiniMax’s own capacity, after we met MiniMax’s vendor verification requirements for accuracy and reliability.
The Inferact Platform serves this agentic workload with guaranteed SLAs and auto-tunes the endpoint against live traffic, while MiniMax keeps control of how its model is served.
Inferact also leads day-0 @vllm_project support for each MiniMax release, so the model serves at production quality the day the weights ship. The engineering from both efforts is upstreamed to vLLM.
Thanks to the MiniMax team for the partnership. Read the full announcement:
inferact.ai/news/minimax-par…
A 社把 moshi 作者的 Claude 账号给封了，他是美国永久居民，只是在中国创业。
A 社乘他续费完毕，而且是 team plan， 一窝端，不退钱，空手套白狼。
这是抢劫吧！
running a US company
being a US permanent resident
using claude only for personal + business work
temporarily staying in China
took my annual renewal for the team plan
2 days later: account is banned
@claudeai, can you at least explain what happened?
and if being physically in China is a hard restriction, then just refund the unused subscription and maybe don’t ban accounts right after taking their annual renewal
用 Opus 5.5 做视频，和做图片一样，先定风格很重要：手绘、3D……选哪种，出来的感觉完全不同。
推荐做知识类视频的推友们，可以看看这个有 5.2k Star 的开源项目：lemo-Opuscar。
最近又新增了 8 种视频风格，我最喜欢「体育战术板」那款，光看风格就很有拆解战术、讲清思路的感觉。
知识博主的选题库要攒，视频风格库也必须攒一个！
github.com/lemomo-ai/lemo-op…
