一只橘子 @Sayed_Dakar

@astigtm2159 这种小跑车其实在琼海那种窄路更灵活，海边大直道反而不显优势。

@ciaren_29 这台车的坐姿高度确实有点像教练车的感觉，第一次坐上去会先松口气。

@Laney_Kuklinski 我之前在吴忠试过一台同类型的，市区基本就两种状态：用电省，踩下去跟得上，没大毛病。

@bondoc_mae 长车头低座舱这套比例本来就少见，黑色反而把线条藏起来了，得凑近才看得出

@cv_endra22_plk 准新车这个状态其实最抢手，车漆和轮毂都没多少磨损，挑的时候重点看下胎压和底盘护板就行。

GPT 6 Astra can get worse when you give it more context 4 GitHub repos that stop a huge context window from turning into token gravity trick: more tokens do not automatically mean more useful information past a certain point the model starts searching through noise instead of reasoning over signal here's the setup: → Repomix packs an entire repository into structured AI ready context github.com/yamadashy/repo… → Serena retrieves and edits code semantically instead of dumping every file into the prompt github.com/oraios/serena → Context7 injects current library documentation only when the agent actually needs it github.com/upstash/contex… → RAGFlow retrieves the relevant evidence before it reaches the model github.com/infiniflow/rag… the architecture is simple: raw context → compress → retrieve → inject only what matters → reason → refresh the goal is not the biggest context window the goal is the smallest context that still contains the answer paste this into Claude Code ↓ "Build a context engineering layer using Repomix Serena Context7 and RAGFlow Use Repomix to create a structured representation of the repository Use Serena for semantic code retrieval so unrelated files stay outside the active context Use Context7 to fetch relevant library documentation only when required Use RAGFlow to retrieve external knowledge before sending context to the model Create a context budget for every request Rank information by relevance Remove duplicate stale and low value context Show exactly what enters the final prompt before execution" bookmark this before dumping another 100k tokens into your agent

Marlen @Marlenuii

2 weeks ago

x.com/i/article/2109…

2 2 49 98K 109

I don’t think people realize how big this Grok Bot update is this is a real game changer for anyone who runs a business, an agency or a creator account on X here are the 5 routines worth setting up first: · the morning scan: 5 things that matter, posts to reply to, one idea to write about · the lead miner: people on X asking for exactly what you sell, with a reply ready · the complaint miner: your competitor’s unhappy users, and whether you solve their problem · the inbox: DMs sorted, every pitch researched before you answer · the weekly report: mentions, sentiment, top posts, what competitors shipped none of them contact anyone. they read, draft and sort. you approve all 5 prompts are in the image, copy and paste ↓

Morlex @0xMorlex

2 weeks ago

Grok Bot just got a serious brain upgrade With Claude Opus 5.5 entering the stack, one Bot could run an entire AI team that researches, writes code, uses tools, verifies results, and delivers finished work Here's how to build it in 10 steps: x.com/i/article/2108…

61 89 860 355K 2K

@k1rallik Nice work

I don’t understand why so few people use Grok Bot inside Blender Same prepared NASA/JPL rover asset. Same assembly brief. Claude Opus 5.5: 36 minutes. $24. Grok Bot: 3 minutes. $2. Watch 00:17. The $2 rover has its arm raised. The other still has its camera mast floating above the chassis. Full prompt in the comments.

cranger @crangerus

4 days ago

Elon Musk's Grok Bot has a cloud computer. What should you actually make it do? Here are 15 prompts for code, research, and writing. Copy one, give your bot a job, and check what comes back. x.com/i/article/2112…

1 4 27 34K 37

@jazzyboobee 顺带说一句，同样的钱拿去打车也就一趟起步，短距离还是它划算

@fireonwheels 想问下保养这块，一年大概花多少？跟油车比差多少你有数吗

@ORenee702 照片看着确实很温柔，但转向‘挺醒’这点我有点好奇，是低速轻还是高速变沉那种？

@livin_red 07年那代我只在山路短距离试过一段，起步之后完全不像停在马路边那个样子，这个感受挺私人，凑合拿去对比。

@chi056 宽而不笨这句话放别人嘴里我是不信的，配图倒还算站得住

@frontm_an this mbalax groove never gets old live

@Felipe60019664 已经转发给正在看小车的朋友了，这种风格应该很对ta胃口。

@tdylbutler 洛阳也能安排的话，大概什么时候去？我在洛阳，想跟着凑个试骑。

this is the most UNFAIR thing on the internet today Jev is the decision layer that turns TradingView alerts into actual trades, no manual chart-watching required i wired it into my own TradingView setup last week, now my alerts execute themselves - charts are cooked here's how to wire Jev into TradingView: 1. set up TradingView alerts on whatever indicators or price levels you already trust - no custom scripting required, just the built-in alert conditions 2. point each alert's webhook at your own Jev endpoint instead of a notification 3. Jev reads the alert payload - symbol, price, which condition fired - and checks it against your rules 4. feed anything ambiguous to Claude Opus 5.5, which reasons through context TradingView's alert alone can't capture 5. Jev returns a yes/no/size decision in under a second, for a fraction of a cent 6. your execution bot only fires on Jev's approved signals, never on the raw alert alone 7. every decision gets logged - alert, verdict, outcome - so you can see exactly which setups Jev actually trusts 8. you end up with a TradingView workflow that filters itself, built on alerts you already pay for i won't say anything more:

bl888m @bl888m_eth

2 weeks ago

Jev might be the fastest decision model you can trade with right now calibrated buy/sell calls in milliseconds I broke down the full HFT framework I built with it, from zero, open-source x.com/i/article/2109…

3 3 31 22K 52

@Mahaximus_ the tamani solo actually carries the whole track, respect

@josie_valino 转给正在纠结买车的朋友了，让他自己去感受下脚下顺不顺