小星聊车日常分享

今天看车最大的收获是知道了什么是HUD抬头显示｜看车日常 @marilyn_1003

[小星聊车日常分享 @bramdelavio]

@OpenNewsGr grind hold, upshots flow hits two-minute save here

[شركة كشف تسربات عزل خزانات بالرياض @5bu]

عزل الأسطح بأفضل تقنيات العزل الحديثة بالرياض 🔥🔧
خدمة فنية عالية الدقة بأفضل الوسائل الحديثة بالرياض 🔥🔧
تواصل معنا من خلال الواتساب
https://t.co/wBVB4zGIJz
#كشف_تسربات_المياه
@كشف@تسربات@المياه
#نظام_المعاملات_التجارية

[小星聊车日常分享 @bramdelavio]

@Noor_daood0 fruity acid gets away early, short way grind on god

[Lana & Hala Family's 🇵🇸🍉 @hazemkhilla]

MY CHILD IS DYING, PLEASE😭

︎
︎

︎

︎

︎
︎
︎

︎

︎

︎

︎don’t forget me🥺
︎

︎

︎

︎

︎

︎
︎

︎
please care

︎
︎

︎
︎

︎
︎

︎

︎

︎

︎
︎

︎

I beg you, I need treatment.

I kiss your hands

Help my sick child, he's dying

https://t.co/RyaPLGSbZv

[小星聊车日常分享 @bramdelavio]

@francescoinweb3 The V60 bloom always smells incredible.

[小星聊车日常分享 @bramdelavio]

@dragon459 @CourtneyLamar 下坡刹车脚感我觉得挺稳的，能量回收介入很线性，不会突然一顿。你是担心重刹点头还是什么？

[小星聊车日常分享 @bramdelavio]

连云港这台LiveWire电摩像科技公司御姐，不靠声音吸引人，右手一动却特别直接。徐州、盐城、宿迁想看电摩骑行感受，主页继续。

[小星聊车日常分享 @bramdelavio]

@MrsRiley09 哈哈第一次都这样，那个推劲回去要缓好几天。你试的是运动模式还是普通模式？两种差挺多的

[小星聊车日常分享 @bramdelavio]

@jstalion_sg 哈哈对，就是那股冷静劲儿，第一下拧下去反而比想的冲。你之前骑的是油车还是踏板？感受会不太一样。

[小星聊车日常分享 @bramdelavio]

@demonacoh 对，老小区这种场景太对了，没声不吵邻居，早上出门也不会被嫌弃。你那边充电位好找吗？这个比声音更影响体验。

[小星聊车日常分享 @bramdelavio]

@shawty_uh_jerkk 我实测市区通勤大概能跑到标称的八成多，走走停停还能接受，连续高架拉高速掉电会明显一些。你关注的是哪个版本？

[小星聊车日常分享 @bramdelavio]

@mao091 充电还算省心，家里普通插座一晚上就差不多了，保养比油车更省事，不用换机油那套。你平时主要在哪边骑？市区通勤和郊区跑法差别还挺大。

[小星聊车日常分享 @bramdelavio]

@patriciabastes 这个确实是个问题，低速行人听不到，我现在路口都习惯减速多看两眼。你是在哪种路况下觉得最明显？

[小星聊车日常分享 @bramdelavio]

@cutiegirl31 车重两百出头，推起来还好，主要是重心低，女生挪车不算费劲。你身高大概多少，我帮你估一下？

[小星聊车日常分享 @bramdelavio]

@rockhey 徐州有门店能约试驾，线下试比线上下单靠谱，具体地址我回头发你？

[小星聊车日常分享 @bramdelavio]

@jenenerald22 安静跑山真的只剩风声和胎噪，注意力全在路线上。你朋友那台是啥车？

[小星聊车日常分享 @bramdelavio]

@CourtneyLamar 对，一拧就走那种直接感真容易上瘾，声音反而成了次要的。你平时骑油车还是电车多？

[小星聊车日常分享 @bramdelavio]

@LogicalThesis @IBKR waited for an evening and the cake came out lopsided

[kozh ./ @mr_kozh]

THIS IS PURE F*CKING TREASURE

10 GitHub repos with 489K+ combined stars to build your own Grok Bot-style AI team

one resource per layer - from the first agent to shared memory, real tools and verified output

none of these repos contains Grok Bot’s source code.

together, they give you the open building blocks behind the same idea:

agents that can divide work, use computers, remember what happened and continue without you manually routing every step.

(1/10) OpenAI Agents SDK

create the agent runtime, handoffs, guardrails, sessions and tracing

▸ https://t.co/YfciOe0OTD

(2/10) CrewAI

turn individual agents into a coordinated crew with responsibilities, shared tasks and controlled execution

▸ https://t.co/lKIOS1xX56

(3/10) LangGraph

make the workflow durable so agents can pause, resume, preserve state and continue after a failure

▸ https://t.co/HyN8rRnPFa

(4/10) MCP Servers

connect the team to files, databases, developer tools and external services through a shared tool protocol

▸ https://t.co/S2hthf3ni8

(5/10) E2B

give every agent an isolated computer where it can execute code, manipulate files and use desktop applications safely

▸ https://t.co/t0UKIQu5Vz

(6/10) Browser Use

let an agent open websites, navigate interfaces, complete forms and perform browser work instead of only describing it

▸ https://t.co/wOd3ee8rXo

(7/10) Mem0

store persistent user and agent memories without resending the entire conversation on every request

▸ https://t.co/ta4a63b40D

(8/10) Graphiti

turn conversations, decisions and changing facts into a temporal knowledge graph the whole agent team can query

▸ https://t.co/9CueTUQlSx

(9/10) Pydantic AI

force agents to return typed, validated outputs instead of handing the next agent another vague paragraph

▸ https://t.co/FQciSs0jj6

(10/10) Phoenix

trace every tool call, inspect failures, run evaluations and see exactly where the team started producing bullshit

▸ https://t.co/RTKicm2t8L

the stack:

agent runtime
→ specialized team
→ durable workflow
→ tools
→ isolated computers
→ browser actions
→ persistent memory
→ shared knowledge
→ validated handoffs
→ traces and evals

the difficult part isn’t spawning ten agents.

it’s making sure agent number ten can trust what agents one through nine already did.

build the coordination layer correctly and your “chatbot” stops waiting for prompts.

it starts operating like a small company.

save this stack before building another AI team that forgets everything after one conversation ↓

[GenAI @0xGenAi]

F*CK. The Grok Bot team built a $2.4M team and put a whole company of bots on the marketplace for free

Nobody has to build them from zero anymore. I mapped 16 of them to 16 seats, one click each

Systems admin, research, prototyping, QA, engineering lead, code audit, integrations, SEO, copy, video, clips, SDR, account management, procurement, PM, recruiting

Import Bot moves your Claude, Codex or ChatGPT setup into Grok Bot

Lingxi Li's Engineer Bot hands coding to cloud agents and watches the PRs. Merges stay with you

Lauren Tan has two bots on the team: tinkabot turns an API into an MCP plugin, dr eggbot builds the bot the marketplace is missing

Haggle Bot finds the SaaS seats you pay for and don't use, then drafts the vendor counteroffer

Price those seats at $100K-$250K each and the payroll comes to $2,415,000 a year

Adding all 16 costs $0

No prompt to write from scratch. You connect your tools and decide what needs your approval

Most people will spend this month building one bot

16 are already built and waiting for a seat

Full guide below ↓ which bots to pick and how to set each one up