Sabitlenmiş Tweet
很好吃
3K posts

很好吃
@ned714
职业:IT (极客) 业余发一些:科技的/运动的/阅读的/电影的/音乐的/个人的
Hangzhou, China Katılım Ekim 2012
342 Takip Edilen151 Takipçiler

@ned714 @o_lalonde @DougDougFood It’s not a joke. Chairman Mao (RIP) has literally dreamed of a white guy getting his KFC chicken sandwich while on a chinese bullet train going 300km/h. Dissapointed you don’t know this
English

@o_lalonde @DougDougFood Perhaps I've been lacking a sense of humor lately.
Some people on X like to discuss ideology-related topics
I really hate it.
English

我记得 Pro 版是为设计师和电影制作人设计的。
如今有了 AI,设计师们不再需要这个了
现在目标用户应该选择大内存版本。
Apple Hub@theapplehub
Apple has officially discontinued the Mac Pro and has no plans for a future model 🚨
中文

@applesclubs I remember that Pro was designed for designers and film production.
Now with AI, designers don't need this anymore
Now the target users should go for the large memory option.
English

@RealXavier011 This kind of video is useless.
Real dust is what sticks to the surface.
Just use your brain a little and think about it
English

you know you made it when you get copycats 😂, this guy literally COPIED 1-1 SlapMac.com. Didnt even bother to change anything. WOW

English

最近在折腾 OpenClaw,发现制约 Agent 进化的往往不是智商,而是数据“鲜度”和“纯度”。
调研了一圈主流的自媒体和网站采集工具,整理出一份我的私藏清单,建议收藏:
1️⃣ 垂直平台“收割机”
MediaCrawler:国内社媒(红/抖/站/快)的全能爬虫,懂的人都懂。
🔗 github.com/NanmiCoder/Med…
Agent-Reach:专门适配 Agent 的社媒数据获取,接入灵活。
🔗 github.com/Panniantong/Ag…
wechat_articles_spider:公众号历史文章采集的最佳实践。
🔗 github.com/klin-h/wechat_…
2️⃣ 浏览器即 API(深度交互)
Web-access:补齐了 Claude Code 联网策略,支持 CDP 浏览器操作。
🔗 github.com/eze-is/web-acc…
BB-browser / OpenCLI:把浏览器变成可编程的 API,适合处理各种反爬和动态加载。
🔗 github.com/epiral/bb-brow…
🔗 github.com/jackwener/open…
3️⃣ LLM 喂料与结构化
r.jina.ai:一行代码网页转 Markdown,Jina 依然是目前体验最好的。
Xcrawl:推荐搭配 OpenClaw 使用,现在注册还送 1000 积分,简单省心。
xcrawl.com/?keyword=dptgb…
Crawl4AI:2026 年最火的开源爬虫,专门为 RAG 和 Agent 设计。
🔗 github.com/unclecode/craw…
4️⃣ 搜索与 MCP 生态
Clawhub:提供的 OpenTwitter 和 OpenNews MCP 插件,是 OpenClaw 的原生搭档。
🔗 clawhub.ai/infra403/opent…
Valyu:Brave 免费额度调整后的首选搜索替代品,质量极高。
5️⃣ 知识闭环
Obsidian Web Clipper:所有采集到的精华,最后都要通过它一键沉淀到 Obsidian,形成自己的“第二大脑”。
🔗 obsidian.md/clipper
💡 总结:对于独立开发和内容创作者,我目前的推荐组合是:OpenClaw + Xcrawl (API) + MediaCrawler (本地任务) + Obsidian (笔记)。
大家最近在用哪些好用的爬虫或者 MCP 插件?欢迎在评论区安利,我再去折腾一下!👇
Yanhua@yanhua1010
养虾的朋友,你们数据源用的什么? 最近我试了 xcrawl,感觉挺好用的。之前 jina 一些轻量场景还行,但反爬强的站就抓不动了。xcrawl 这块做得比较扎实,JS 渲染、验证码都能搞定,输出的 Markdown 也干净,直接喂给 Agent 没问题。 目前已有4 个现成的 Skill:抓单页、爬全站、搜索、站点地图,赶紧给你的agent装上:github.com/xcrawl-api/xcr… 新用户送 1000 积分,够玩一阵了,从这进入👇xcrawl.com/?keyword=dptgb…
中文

@elonmusk @AdamLowisz It seems like all of humanity's nightmares could be AI-generated. lol.😆😆😅😅
English

@AdamLowisz I’ve been having a lot of AI nightmares lately :(
English

Elon, sit back, enjoy the singularity, and hope for the best 😌
Elon Musk@elonmusk
@Rothmus That is right about the time that AI exceeds all human intelligence
English

@TansuYegen The timbre somewhat resembles that of a church organ. How does this glass produce sound?
English

@chinafutureclub Those who love mechanical movement all hate electric vehicles.
English

@SundaeDivine "The Art of War" in the chapter "Offensive Strategy" states: "To subdue the enemy without fighting is the acme of skill. Next best is to disrupt his alliances. The next, to attack his army. The worst policy is to attack cities." Reading good books broadens one's horizons.
English



















