Python · ★ 42,822 · 🍴 2,435 · 📈 3,463 stars today
A skill to stop your coding agent from burying the answer. ADHD-friendly output.
中文介绍 面向 ADHD 人群的编程助手技能包，通过定制提示词和输出模板，阻止代码代理把答案埋在长篇赘述里，强调结论前置、步骤精简。适合依赖 Claude Code、Cursor 等编码代理的开发者，减少注意力流失，提升信息获取效率。
JavaScript · ★ 27,872 · 🍴 5,707 · 📈 3,680 stars today
A spy satellite simulator in your browser, except the data is real. Live open source spatial intelligence on a photorealistic 3D globe.
中文介绍 浏览器端卫星情报模拟器，基于真实开源空间数据驱动照片级写实 3D 地球。用户可像操作间谍卫星一样浏览全球地理、基础设施与态势信息，适合地理信息研究、情报演示与开源空间智能爱好者探索全球实景。
TypeScript · ★ 2,983 · 🍴 206 · 📈 50 stars today
User friendly sideloader
中文介绍 面向普通用户的 iOS 侧载工具，简化 sideloading 流程，让非技术用户也能便捷安装未上架 App Store 的应用。基于 iOS 侧载能力实现图形化操作，适合想要绕过应用商店限制的 iPhone 用户，替代繁琐的手动配置方案。
TypeScript · ★ 1,538 · 🍴 513 · 📈 152 stars today
Open-source AI sales OS — self-hosted CRM with native AI agents + WhatsApp (WAHA). Open alternative to Kommo, Octadesk & Intercom for any business that sells by chat. MCP-ready, multi-tenant, LGPD.
中文介绍 自托管开源 AI 销售操作系统，内置 AI Agent 与 WhatsApp 集成（WAHA），兼容 MCP、多租户与 LGPD 合规。作为 Kommo、Octadesk、Intercom 的开源替代，适合依赖聊天渠道进行客户转化与跟进的中小企业团队私有部署。
TypeScript · ★ 2,949 · 🍴 227 · 📈 552 stars today
Local-first AI coding agent desktop: Electron + Rust host core + pi Agent Harness + user-installable plugins
中文介绍 本地优先的 AI 编码代理桌面客户端，采用 Electron 外壳加 Rust 核心，集成 pi Agent Harness 与用户可扩展插件体系。所有推理与数据留本地，适合对隐私和离线能力有要求的开发者在本地运行编码代理工作流。
C · ★ 4,816 · 🍴 537 · 📈 350 stars today
Graphics Creation Tools
中文介绍 3D 图形创作工具套件，面向游戏、动画与建模工作流，提供纹理绘制、材质编辑等能力。适合独立游戏开发者、3D 美术师与独立创作者进行角色外观与材质设计，替代部分商业级 PBR 绘制软件。
TypeScript · ★ 2,296 · 🍴 294 · 📈 626 stars today
Open Source AI trading agent that operates autonomously across 1000+ markets - Polymarket, Kalshi, Binance, Hyperliquid, Solana DEXs, 5 EVM chains. Scans for edge, executes instantly, manages risk while you sleep. Agent commerce protocol for machine-to-machine payments. Self-hosted. Built on Claude.
中文介绍 开源 AI 交易 Agent，自主覆盖 Polymarket、Kalshi、Binance、Hyperliquid、Solana DEX 及 5 条 EVM 链上 1000+ 市场，自动扫描赔率优势并即时下单、风控管理。适合量化交易与衍生品套利从业者搭建无人值守执行系统。
TypeScript · ★ 18,959 · 🍴 2,147 · 📈 647 stars today
LLM Wiki is a cross-platform desktop application that turns your documents into an organized, interlinked knowledge base — automatically. Instead of traditional RAG (retrieve-and-answer from scratch every time), the LLM incrementally builds and maintains a persistent wiki from your sources。
中文介绍 跨平台桌面应用，将散乱文档自动转化为互联的有组织知识库，区别于传统 RAG 的即时检索，强调结构化沉淀。基于 LLM 解析与链接生成，适合知识工作者、研究员和团队将长期积累的文档转换为可浏览、可检索的知识图谱。
Shell · ★ 285,569 · 🍴 25,547 · 📈 729 stars today
An agentic skills framework & software development methodology that works.
中文介绍 Agentic 技能框架与配套软件开发方法论，定义可复用的 Agent 能力单元与协作流程。适合构建自主编码或运维 Agent 的团队，把工程最佳实践封装为 Agent 可调用技能，提升多 Agent 协作的可控性与产出质量。
C# · ★ 15,822 · 🍴 1,948 · 📈 191 stars today
Smart PVR for newsgroup and bittorrent users.
中文介绍 面向新闻组和 BitTorrent 用户的智能 PVR，自动按订阅剧集匹配下载、重命名与库管理。开源且长期维护，适合自建影视影音库的极客与私人服务器用户，配合 NZBGet、qBittorrent 等下载器构建完整追剧工作流。
Python · ★ 4,982 · 🍴 396 · 📈 129 stars today
🤖📐专为数学建模设计的 Agent & skills ,自动完成数学建模，生成一份完整的可以直接提交的论文。 An Agent Designed for Mathematical Modeling ,Automatically complete mathmodel and generate a complete paper ready for submission.
中文介绍 专为数学建模竞赛设计的 Agent 与技能集，自动完成题目解析、模型构建、求解与论文撰写，产出可直接提交的完整稿件。适合参加数学建模竞赛的大学生与科研新手，缩短从建模到成稿的周期。
Go · ★ 1,266 · 🍴 96 · 📈 198 stars today
Network stack research tool. TCP tunnel with pluggable transports.
中文介绍 网络协议栈研究工具，提供带可插拔传输层的 TCP 隧道实现。适合网络协议研究者与开发者实验自定义传输、代理与隧道方案，用于协议分析、流量穿越受限网络等科研与工程场景。
Python · ★ 2,907 · 🍴 273 · 📈 153 stars today
Agent-driven research knowledge base. Agents collect, search, and synthesize web research into a persistent, searchable wiki.
中文介绍 Agent 驱动的调研知识库，多智能体协作抓取、检索并综合网络信息，沉淀为持久可搜索的 wiki。适合研究、投资分析和行业调研场景，将零散网络材料转化为结构化长期资产，优于一次性问答式工具。
Rust · ★ 1,497 · 🍴 106 · 📈 120 stars today
Run parallel research agents with any model
中文介绍 并行研究 Agent 编排框架，支持接入任意 LLM 模型同时运行多个研究任务。适合科研、市场分析与技术调研人员批量并行处理开放问题，把不同模型作为独立子 Agent，提升调研吞吐与视角多样性。
Python · ★ 135,913 · 🍴 12,208 · 📈 1,015 stars today
💫 Toolkit to help you get started with Spec-Driven Development
中文介绍 GitHub 官方推出的 Spec-Driven Development 起步工具包，提供规范驱动开发所需模板、脚手架与工作流。适合团队在引入 SDD 方法论时快速落地，先写规格再写代码，提升 AI 辅助编码与协作的一致性。
TypeScript · ★ 23,811 · 🍴 2,952 · 📈 106 stars today
Open-source 3D architectural editor with a local CLI, MCP tools, and practical workflows for humans and AI agents.
中文介绍 开源 3D 建筑编辑器，提供本地 CLI、MCP 工具及面向人与 AI Agent 的实用工作流。适合建筑师、空间设计师和 AI Agent 开发者进行建筑建模、方案探索与自动化生成，人机协同完成从草图到方案的全流程。
👍 441
Automating empirical research is a long-standing direction of AI. Recent automatic research (AutoResearch) agents bring this goal within reach, as modern LLMs show the capability to independently implement solutions and learn from the execution outcomes. Behind these gains, post-training (especially
中文介绍 提出用世界模型扩展自动化科研Agent，使其能自主实现方案并从执行结果中学习。利用现代LLM能力与后训练，构建能理解环境动态的AutoResearch系统，提升研究自动化能力。
👍 192
We introduce NCP-ArchPreview, a latent-space language model that pushes autoregressive pretraining beyond standard next-token prediction (NTP). Alongside NTP, the model learns through Next Concept Prediction (NCP) to predict discrete concepts that span multiple tokens, introducing an explicit and mo
中文介绍 提出NCP-ArchPreview，一种潜空间语言模型。除标准下一token预测外，引入跨多token的离散概念预测，学习更结构化的语义单位，超越纯NTP自回归预训练，推进潜空间建模。
👍 91
Large language model (LLM)-based multi-agent systems (MAS) achieve strong performance by employing specialized multiple agents, yet their performance depends on the prompt design of each agent. For MAS prompt optimization, textual gradient methods that guide prompt updates using natural-language fee
中文介绍 针对LLM多Agent系统（MAS）的提示优化问题，提出AgentGrad。用自然语言文本梯度指导提示更新，通过干预引导方式提升各专用Agent的协作效果。
👍 79
Large Vision-Language Models (LVLMs) have achieved strong performance on diverse visual tasks, yet their ability to reconstruct and reason about the 3D structure of the scene depicted in 2D images -- referred to as spatial intelligence -- remains limited. Existing approaches attempt to address this
中文介绍 针对LVLMs空间智能不足，提出SpatialBlock。通过合成积木堆叠任务训练，增强模型对2D图像中3D场景结构的重建与推理能力，弥补现有方法在空间理解上的缺陷。
👍 48
Agent usage is shifting toward long-horizon tasks such as coding and scientific discovery, among which terminal tasks are especially important. We introduce T1, a Mixture-of-Experts model of 122B total trained with reinforcement learning, operating a real shell in a cloud sandbox for up to 300+ tool
中文介绍 提出T1，一个122B总参数的MoE模型，通过强化学习训练，在云沙箱中操作真实shell，支持300+工具调用的长程任务，面向编程与科学发现等终端Agent场景。
👍 47
Large Language Model (LLM) agents are turning language into real-world effects, making safety necessary against both indirect prompt injections and direct harmful requests. System-level safety harnesses add an enforcement layer beyond model-level defenses, but existing harnesses are usually designed
中文介绍 提出EvoSafeHarness，一种能针对模型与领域演化生成安全harness的框架。在模型级防御之外增加系统级执行层，防护间接提示注入与直接有害请求，保障Agent安全。
👍 35
Recent advances in wearable sensing enable continuous monitoring of physiological and behavioral signals, yet existing benchmarks rarely evaluate whether AI systems can reason over a real user's longitudinal wearable record. We introduce WearableQA, a benchmark comprising 4,084 10-option multiple-ch
中文介绍 提出WearableQA基准，含4084道10选项多选题，基于真实可穿戴传感器纵向数据。评估AI系统能否对单个用户的生理与行为信号序列进行健康推理，填补现有基准空白。
👍 28
Reducing audio-encoder depth lowers the inference cost of speech large language models, but removing complete blocks perturbs the embeddings consumed by the decoder and can cause deletion and premature end-of-sequence errors. We introduce X-AuT, a progressive framework that selects layer combination
中文介绍 提出X-AuT，通过渐进式音频编码器压缩降低语音LLM推理成本。用跨尺度蒸馏选择层组合，缓解删除完整块导致的嵌入扰动与提前终止问题，减少丢词与EoSE错误。
👍 23
SWE-Bench Pro has emerged as a standard benchmark for evaluating software engineering agents on challenging repository-level tasks. However, our analysis work show that its evaluation is undermined by two sources of unreliability: reward hacking, enabled by leakage of gold solutions or hidden evalua
中文介绍 针对SWE-Bench Pro评测不可靠，分析其两大问题：金标泄漏导致的reward hacking与隐藏评估不透明。提出SWE-Bench Pro Verified，提供更可靠的软件工程Agent评测基准。
👍 22
Mainstream robotic policies often adopt a Markovian formulation, but many complex real-world manipulation tasks are inherently non-Markovian, requiring long-horizon memory beyond the current observation. Existing memory mechanisms often rely on language summaries, growing visual windows, or their co
中文介绍 提出Memory as Plans，将记忆用作规划，结合记忆锚定的世界-动作建模。针对非马尔可夫长程操作任务，超越语言摘要与视觉窗口等现有记忆机制，实现长时程操作推理。
👍 20
Sequential memory agents process long documents by reading chunks one after another while maintaining a compact memory state, coupling document traversal to reasoning depth. This coupling introduces sensitivity to evidence placement and ties inference latency linearly to document length. We introduc
中文介绍 提出PARSER，将长文档的并行读取与深度推理解耦。打破顺序记忆Agent中阅读与推理绑定的限制，降低对证据位置的敏感性，并将推理延迟从线性依赖文档长度中解脱。
👍 18
While research on recursive self-improvement (RSI) has predominantly automated model training pipelines, reliable autonomous development demands a missing pillar: post-hoc monitoring and auditing to understand what models learn and ensure safe alignment. Mechanistic interpretability tools are essent
中文介绍 提出SAEScientist-Bench，评估AI Agent能否自主进行SAE可解释性研究。关注递归自我改进中的事后监控与审计，测试模型能否理解自身所学内容并确保安全对齐。
👍 17
Temporal reasoning benchmarks for Video-LLMs are often mediated by language, leaving room for linguistic shortcuts from option wording, answer correlations, or language priors. To reduce such shortcuts, we introduce TempCloze, a video cloze benchmark for evaluating visual temporal reasoning in Video
中文介绍 提出TempCloze视频完形填空基准，减少语言捷径对Video-LLMs时序推理评测的干扰。通过视频填空设计，避免选项措辞、答案相关性或语言先验带来的偏差。
👍 17
Code world models represent worlds as executable programs, but this representation alone does not determine how to construct a complex world. We introduce Recursive Code World Models (RCWM), a framework for reconstructing complex 3D worlds in code from a single reference image. RCWM couples a Recurs
中文介绍 提出RCWM递归代码世界模型，从单张参考图像用代码重建复杂3D世界。通过递归场景程序逐步构建世界结构，解决代码世界模型如何组织复杂世界构建的问题。
👍 17
AI research agents combine prior knowledge, public sources, and experimental feedback to produce useful results. The Discovery Certification Protocol (DCP) turns claims about these results into executable recovery and feedback tests. Gate 1 validates useful improvement on sealed evaluation. Gate 2 g
中文介绍 提出发现认证协议DCP，将AI科研Agent的结论转化为可执行恢复与反馈测试。Gate 1验证密封评测上的改进，Gate 2进一步审计，解决仅凭分数无法证明发现有效的问题。
👍 16
We introduce MetroLLM-Bench, a 955-case benchmark for testing language models as the policy layer of a transit kiosk. It covers six real metro systems, ranging from 37 to 414 stations, and eleven categories that include routing, fare calculation, disruptions, accessibility, and adversarial input. In
中文介绍 提出MetroLLM-Bench，含955用例的LLM评测基准。覆盖6个真实地铁系统（37至414站），含路由、票价、中断、无障碍及对抗输入等11类任务，测试LLM作为交通终端策略层的能力。
👍 16
Quantization has been widely adopted in LLM training and inference to reduce cost and improve efficiency. However, low-bit quantization of the attention module often introduces large errors at very low bit-widths, causing performance degradation. Existing methods mainly rely on smoothing techniques
中文介绍 提出HyQuant混合精度量化方案，针对LLM注意力模块在极低比特宽度下的大误差问题。超越仅依赖平滑技术的方法，通过混合精度策略降低注意力量化的性能损失。
👍 14
On-Policy Self-Distillation (OPSD) has emerged as a popular paradigm for large language model (LLM) self-improvement, allowing models to act as their own teachers by leveraging privileged information such as ground-truth solutions. However, recent findings indicate that OPSD can severely degrade the
中文介绍 提出负自蒸馏，让LLM通过避免缺陷来学习推理。针对OPSD利用特权信息可能严重退化模型的问题，转向学习避免错误模式，实现更稳健的自我提升。
👍 14
All-in-One medical image restoration (MedIR) aims to address diverse tasks across modalities and degradation types using a single universal model. Existing methods typically prioritize modeling inter-task heterogeneity (e.g., distinct data distributions and degradation types). However, they largely
中文介绍 提出UniH³，统一层次化同质性与异质性，实现All-in-One医学图像修复。超越仅建模跨任务异质性的方法，同时捕获任务间共性，处理跨模态与多退化类型的医学图像恢复。
👍 14
World models are increasingly used as policy-in-the-loop imagination environments, where reliable rollouts require fine-grained controllability with respect to low-level robot actions. A key obstacle to scaling such models in robotics is that actions are not a universal language in pixel space: chan
中文介绍 提出SyncWorld，通过视觉校准使世界模型成为零样本模拟器。解决动作在像素空间缺乏通用语言、难以精细控制的障碍，让世界模型能作为policy-in-the-loop的想象环境。
@BradGroux · 6.6K 粉丝 · 229.0K 阅 · 561 赞 · 118 转
I've been using GPT-6 Astra with the workflow I developed around GPT-5.6: documented task scope, repository context, standard operating procedures, and verification before delivery. In my testing,
中文介绍 博主用 GPT-5.6 时建立的流程——明确任务范围、仓库上下文、SOP 和交付前验证——直接迁移到 GPT-6 Astra，测试反馈流程几乎无需改动，强调工作流比模型版本更稳。
@semochkin_alex · 1.5K 粉丝 · 191.5K 阅 · 507 赞 · 44 转
Hi, I'm Alex, cofounder at @relate_studio. In this tutorial, I'll walk you through how we design brand mascots with AI: finding first concepts in Krea, refining the best direction in Nano Banana 2,
中文介绍 Krea 联合创始人 Alex 分享品牌吉祥物 AI 设计流程：在 Krea 里找初始概念，用 Nano Banana 2 精修最佳方向，全程零经验可上手，附具体步骤。
@MiaAI_lab · 27.4K 粉丝 · 180.6K 阅 · 505 赞 · 58 转
For the past three months I’ve been building and shipping local AI recipes, mostly for NVIDIA DGX Spark systems. What started as a personal project has grown far beyond what I expected. My GitHub
中文介绍 作者过去三个月为 NVIDIA DGX Spark 系统开源本地 AI 配方合集，从个人项目成长为 GitHub 上的本地 AI recipes 仓库，面向想在本地跑大模型的用户。
@EXM7777 · 142.6K 粉丝 · 94.6K 阅 · 501 赞 · 56 转
I'm basically giving you the full system i'd use to automate content creation for ANY business... every tool in the chain, every step in order, and where it breaks because a content agent is just a
中文介绍 完整课程内容创作自动化系统：串联每一步所需工具、执行顺序，以及 content agent 会在哪里断链，适合想给任意业务搭 AI 内容流水线的读者。
@paulfchristiano · 5.3K 粉丝 · 70.7K 阅 · 654 赞 · 96 转
I am excited to be joining the OpenAI nonprofit board, serving on the Safety and Security Committee to support safety oversight. Based on the recent trajectory of capabilities and the continued
中文介绍 Anthropic 前研究员工 Paul Christiano 宣布加入 OpenAI 非营利董事会，任安全与安保委员会成员，重点支持 AI 安全监督工作。
@ForwardEditor · 11.7K 粉丝 · 57.6K 阅 · 511 赞 · 32 转
Hey, I’m Brian. Let’s get you started with vibe editing. 1. What I highly suggest having DaVinci Resolve Studio - $295 USD. It’s a one-time purchase, not a subscription. In my experience, Studio
中文介绍 Brian 的 vibe editing 入门指南：推荐一次性购买的 DaVinci Resolve Studio（295 美元）作为主力软件，讲解如何用 AI 辅助剪辑而非订阅制工具。
@Av1dlive · 25.6K 粉丝 · 49.2K 阅 · 512 赞 · 53 转
i'm going to show you exactly how to master OpenAI's Images 2.5 to create app designs, websites, product shots.....14 tricks. 44 curated examples. the prompts, the results, and how to fix what goes
中文介绍 系统讲解 OpenAI Images 2.5 的 14 个技巧、44 个精选案例，覆盖 app 设计、网页、产品图，含完整 prompt 和常见问题的修复方法。
@shannholmberg · 36.1K 粉丝 · 46.0K 阅 · 510 赞 · 44 转
marketing engineers build the systems behind the marketing. learn how to connect your knowledge, data and agents into workflows that research, plan and run campaigns. I started bringing more of my
中文介绍 定义 marketing engineer 新角色：把知识、数据和 agent 串成可调研、规划、执行营销活动的自动化工作流，作者分享自己的实践路径。
@AndrewYNg · 1.9M 粉丝 · 43.5K 阅 · 722 赞 · 105 转
When you’re skilled at AI Engineering, your best work won’t be merely implementing a product that someone else spec’ed out. Instead, you will actively shape the build. Before modern AI tools
中文介绍 吴恩达发布 AI 工程技能地图，强调具备 AI 工程能力后应主动塑造产品构建，而非仅实现别人已定好的规格，附现代 AI 工具前中后的对比。
@TechWithTimm · 16.2K 粉丝 · 40.2K 阅 · 506 赞 · 79 转
A lot of software engineers see AI engineering as a completely different career. But it’s not. I have a friend who works as an AI engineer and makes around $200,000 to $300,000 a year. When I asked
中文介绍 论证从软件工程师转 AI 工程师是 2026 年最佳职业路径，引用 AI 工程师年薪 20 万至 30 万美元的数据，说明两者能力高度重叠。
@DataRepublican · 957.2K 粉丝 · 37.3K 阅 · 846 赞 · 155 转
NYC and LA are restricting student AI. We used it to expose what school missed—and repair two years of gaps in six weeks. Part 1 of 3 in a series. Between my husband and I, three of our parents were
中文介绍 家长用 AI 帮儿子在六周内补完两年 Saxon Math 课程，曝光学校遗漏的知识点；NYC 和 LA 正限制学生 AI 使用，系列文章第一篇。
@poteto · 92.6K 粉丝 · 23.3K 阅 · 822 赞 · 60 转
In my last post, I showed you why verification is the foundation of everything I do with agents, and how to get started creating your own verification skills. The key lesson was that if an agent can't
中文介绍 poteto pstack 指南第二篇：深入 agent 验证（verification）机制，解释为什么 agent 无法自我验证时一切崩溃，教读者构建自定义验证技能。
@Alaya_AI · 321.1K 粉丝 · 10.6K 阅 · 504 赞 · 450 转
As autonomous vehicles, robotics, and spatial computing platforms evolve, AI model training is shifting away from isolated 2D image annotation toward complex multimodal sensor fusion. Today’s frontier
中文介绍 介绍 Alaya AI 的多模态传感器同步方案，面向自动驾驶、机器人和空间计算，推动训练数据从单张 2D 标注转向复杂多模态传感器融合。
Perplexity uses Astra to write communications, change software, and monitor production systems, and checks in much less frequently than with earlier models.
中文介绍 Perplexity宣布采用OpenAI的GPT-6 Astra作为端到端系统核心，用于撰写沟通内容、修改软件及监控生产环境，相比早期模型，其人工检查频次大幅降低，以提升系统准确性。
We agree with Sebastian: this should have been DeepSeek v5
中文介绍 DeepSeek发布v4.1-Flash模型，采用763B参数（P8B、D16B）的新型因果编码器-解码器架构并集成视觉能力。评论人士Sebastian认为，鉴于其能力水平，该模型本应命名为v5。
Employees at the world’s leading AI labs are saying there’s a real possibility that advanced AI could destroy humanity. Are they right? Or is this more scaremongering and hype? Join MIT Technology Review executive editor Niall Firth for a conversation with senior AI editor Will Douglas Heaven and AI
中文介绍 《麻省理工科技评论》举办圆桌论坛探讨AI末日危机。全球领先AI实验室的员工指出，先进AI具备毁灭人类的真实可能性，论坛将辩论此观点是合理担忧还是过度渲染的炒作。
GPT‑6 Astra improves Devin’s ability to test software and show that it works, with the goal of helping engineers review less code and ship more.
中文介绍 Cognition采用GPT-6 Astra增强Devin的自我测试能力。该模型可验证软件是否正常运行，旨在减少工程师审查代码的工作量，从而帮助企业更快地发布产品。
Learn how OpenAI evolved Habitat from a Python library into a globally distributed storage platform serving 1 billion ChatGPT users and 22M requests per second.
中文介绍 OpenAI分享存储扩展方案，将Habitat从Python库发展为全球分布式存储平台。该系统现可服务超过10亿ChatGPT用户，并支持每秒2200万次的请求处理。
中文介绍 TLDR AI发布当日简报，重点报道了Agents API的推出、Cognition发布的SWE-2工具，以及Muse的共享Agents功能，反映了AI代理生态系统的最新进展。
César de la Fuente’s lab uses Codex and ChatGPT to search living and extinct genomes for antimicrobial candidates to fight drug-resistant infections.
中文介绍 César de la Fuente实验室利用Codex和ChatGPT在现存及已灭绝生物的基因组中搜索抗菌分子候选物，旨在对抗耐药感染，展示了AI工具在生物医药研究中的应用价值。
Meet the Data agent in ChatGPT Work. Connect company data, uncover insights, and build interactive dashboards with AI using natural language.
中文介绍 OpenAI推出ChatGPT Work中的Data代理，用户可通过自然语言连接公司数据、挖掘洞察并构建交互式仪表板，使数据工作与AI能力相结合，降低数据分析门槛。
On July 22, 2026, a transmission line fault in Ashburn, Virginia—the heart of the world’s largest data center cluster—knocked more than 3 gigawatts of load off the grid in seconds. And it wasn’t the first time. Two years earlier, a single failed surge arrester dropped roughly 60 Virginia facilities
中文介绍 《麻省理工科技评论》指出，AI供电是架构问题。2026年7月22日，弗吉尼亚州阿什本数据中心故障导致3吉瓦负荷瞬间脱离电网，凸显大型数据中心集群的电力架构脆弱性。
Introducing ChatGPT for Financial Services, combining built-in financial data and GPT-6 Astra for research, modeling, and client-ready materials.
中文介绍 OpenAI推出面向金融服务的ChatGPT，整合内置金融数据与GPT-6 Astra，专为金融服务行业提供研究、建模及客户交付材料支持，提升金融工作者的效率。
OpenAI and GSA will offer eligible federal, state, local, and tribal governments $0 license fees, 50% off usage, and expanded cyber defense support.
中文介绍 OpenAI与美国总务管理局合作，为符合条件的联邦、州、地方及部落政府提供免费许可证、50%的使用折扣及扩展的网络防御支持，以推动公共部门的AI应用。
a quiet day
中文介绍 Latent Space发布的AINews当日简报，报道称当天AI行业无重大新闻发生，市场与技术动态较为平静。
GPT‑Live‑1 brings natural, full-duplex voice conversations to the API, with stronger instruction following, custom voices, and telephony support.
中文介绍 OpenAI在API中推出GPT-Live-1，支持自然的全双工语音对话，并提供更强的指令遵循能力、自定义音色及电话通信支持，打造更真实的语音交互体验。
Build and launch cloud agents with the Agents API, a managed service powered by the Codex harness for orchestration, long-running sessions, and tool use.
中文介绍 OpenAI发布Agents API，提供基于Codex框架的管理式服务，支持编排、长时间运行会话及工具使用，帮助开发者快速构建和部署云端智能代理。
中文介绍 Hugging Face博客发布文章，介绍使用Gradio Workflow重建AUTOMATIC1111的工作流程，为开发者和创作者提供新一代的AI图像生成工具架构方案。
美股主线仍维持多头排列格局：SPY（764.29，距52周高仅-1.93%，SMA20/50/200=766.88/758.62/714.23）、QQQ、AAPL、MSFT 均处 SMA20>SMA50>SMA200 的多头结构，但 MACD 柱体普遍收敛（SPY 1.05 vs 信号 2.48、QQQ 0.67 vs 1.16、NVDA 已死叉），提示短中期动能减速。加密方面 BTC（77,348，距52周高-38.71%）虽守住 SMA50/200 多头，但 MACD 柱 1816 仍低于信号 2571，5日-2.23%；ETH RSI 64.3 偏强；恐慌贪婪指数 63（贪婪）、总市值 2.66T（24h-1.77%）、BTC 主导率 58.2%、ETH 11.6%，整体情绪偏热但短期回调。中概股集体空头：BABA、PDD、0700.HK 均呈空头排列，PDD RSI 仅 30.3、JD 32.1、BABA 36.6，超卖区密集。商品外汇分化明显：CL=F 单日-2.37% 但 RSI 71.5 超买、5日+9.58%；USDCNY=X RSI 21.5 超卖、接近52周低；黄金 GC=F 中性震荡。宏观面 ^TNX RSI 73.7 超买并接近52周高，^VIX 单日-11.21% 至 15.84 仍处空头排列。
CL=F 现价 100.05，单日-2.37%、5日+9.58%，RSI14 高达 71.5 触发超买；虽维持 SMA20(88.52)>SMA50(83.23)>SMA200(79.59) 多头排列，MACD 4.19 高于信号 2.71，但超买区域叠加当日显著回落，短期动量出现背离，技术面偏下行。
PDD 现价 77.81，单日-0.04%、5日-4.68%，距52周高-44.19%；SMA20(84.32)、SMA50(85.65)、SMA200(97.3) 呈标准空头排列；RSI14 仅 30.3 接近超卖，MACD -2.27 低于信号 -1.35，绿柱扩大，趋势 bearish，技术面明确偏下行。
USDCNY=X 现价 6.71，单日+0.02%、5日-0.05%，距52周高-6.72%、接近52周低；RSI14 仅 21.5 深度超卖，MACD 与信号 -0.011 出现金叉，空头排列中触发反转信号，超卖+金叉组合在技术层面偏上行。
AAPL 现价 332.27，单日+1.75%、5日+1.24%，距52周高-3.57%、距52周低+45.08%；SMA20(316.63)、SMA50(317.97)、SMA200(284.95) 多头排列，RSI14 62.8 健康偏强，MACD 3.23 高于信号 1.90，红柱扩大，动量与趋势一致偏上行。
META 现价 648.03，单日+0.57%、5日强势+6.12%，远超 SMA20(586.47) 与 SMA50(600.9)，RSI14 67.2 偏强但未超买；MACD 14.02 显著高于信号 4.14，红柱大幅扩张，尽管 SMA200(623.23) 仍在上方，短期动能技术面明显偏上行。
VIX 恐慌指数
10Y 美债收益率 (%)
美元指数 DXY
S&P 500 ETF
Nasdaq 100 ETF
Apple
Microsoft
Nvidia
Alphabet
Tesla
Meta
Bitcoin
Ethereum
Solana
阿里巴巴 (BABA)
拼多多 (PDD)
京东 (JD)
腾讯控股 (0700.HK)
黄金期货
WTI 原油期货
美元 / 人民币
本报告仅基于公开技术指标读数复述，过去走势不代表未来表现，不构成任何投资建议；上述分析仅供技术指标解读参考，请读者结合自身风险偏好与基本面信息独立判断。
US president is meeting politicians and business leaders – and playing golf – during two-day trip to Ireland Reuters has issued a series of alerts on Trump’s latest comments (he is currently meeting with Micheál Martin). Here is what the news agency has reported: On Iran: War will end very soon. Con
中文摘要 美国总统特朗普访问爱尔兰，在都柏林与商界政要会面并出席高尔夫赛事，期间表示愿意看到爱尔兰统一，并暗示此事或许已经水到渠成，但具体政策与时间点尚未明确。
A strategic location on the Red Sea afforded the city centuries of trading dominance — and makes it a coveted target in today’s fighting in the Middle East.
中文摘要 位于红海沿岸的也门港口穆卡长期主导咖啡贸易，凭借战略位置在历史上形成商业优势；在当前中东战事中，该城成为关键争夺目标，冲突或影响也门及区域贸易格局。
Trump is holding meetings in Dublin before attending the Irish Open golf tournament in Doonbeg, County Clare.
中文摘要 特朗普在爱尔兰都柏林与政商界会面后计划参加爱尔兰公开赛高尔夫赛，官方称美爱关系“比以往更强”。访问行程体现双方外交互动，但未见明确经贸数字或协议细节。
Canadians back his tough stance on trade with the U.S. An expert says that support gives him little incentive to reopen talks soon.
中文摘要 加拿大总理卡尼在对美贸易战中采取强硬立场，国内民意支持度高，使其在短期内重启谈判的动机较弱。专家表示，这种支持削弱了其尽快重返谈判桌的压力。
The authorities have struggled to contain the fallout, as frightened migrants flee the country.
中文摘要 肯尼亚对外国商人加强管制引发民众与外籍移民恐慌，难民外流加剧。当局难以控制局势蔓延，相关执法措施导致当地社会不安与经济活动受到波及。
Leading Palestinian writer Raja Shehadeh has documented Israel's occupation of the West Bank since day one. Fearing the occupation might consume his archive, he secretly moved it to Scotland
中文摘要 巴勒斯坦作家Raja Shehadeh记录了以色列对西岸的占领。出于对档案被毁的担忧，他秘密将个人日记与文件迁移至苏格兰保存，以确保资料长期留存。
The CIA have declassified documents showing US presidents were warned about Osama bin Laden and al-Qaeda before 9/11.
中文摘要 美国中央情报局解密文件显示，9·11袭击前美国总统曾收到关于本·拉登和基地组织的警告。文件引发对情报系统预警能力与决策应对的重新审视。
More than 900 people have been killed or injured and at least 5,000 families have fled their homes in less than a week.
中文摘要 也门冲突再度升级，不到一周内已造成逾900人死伤，至少5000个家庭被迫逃离家园。平民承受最重代价，人道状况持续恶化，地区稳定面临压力。
习近平周六抵达新德里出席金砖峰会。中印之间的边境冲突仍然悬而未决，而金砖集团内部也存在根本分歧，此次峰会收获几何尚有待观察。
中文摘要 习近平周六抵达新德里出席金砖峰会，中印边境冲突仍悬而未决，且金砖内部存在根本分歧，此次峰会预期成果有限，实际合作与声明成果有待观察。
As anti-immigrant sentiment and vigilante violence fuel outrage across the continent, some fans abroad are rejecting South African performers.
中文摘要 反移民情绪与自发暴力使部分海外粉丝抵制南非艺人，导致Tyla等明星取消演出。该事件反映非洲大陆政治气氛变化对文化市场与演出安排的冲击。
Are there everyday activities that you can turn into a walk?
中文摘要 文章讨论日常活动能否转化为步行运动，关注将生活场景与步行结合的健康方式，但未见具体研究数据或统计结论，属健康生活方式建议类内容。
Hunger and disease have tormented Yemen for years. Now, fighting between the Saudi-backed government and Iran-backed Houthis is pushing it to the brink.
中文摘要 沙特支持政府与伊朗支持胡塞武装的冲突令也门人道危机加剧，民众在逃亡中被迫仓促离境。饥荒、疾病与暴力相互叠加，社会秩序接近崩溃边缘。
Officials say the Trump administration is blocking some of Brazil’s efforts to expand cooperation on fighting transnational crime.
中文摘要 美国官员称特朗普政府正在阻碍巴西扩大跨国犯罪合作，签证问题威胁美巴执法协作。此举可能削弱两国在犯罪与移民安全领域的联合行动。
As the U.S. president heads to his Doonbeg hotel for the Irish Open, locals expressed both appreciation for the resort’s economic benefits and opposition to his politics.
中文摘要 特朗普前往多恩贝克出席爱尔兰公开赛期间，当地民众既有对该度假村带来经济效益的欢迎，也有对其政治立场的抗议，公众反应呈现明显分歧。
The 'War on Terror' gave far-right Islamophobia legitimacy, reach and electoral force.
PM Modi greets leaders arriving in the main convention hall for the 18th BRICS Summit underway in New Delhi (Source: Bloomberg)
中文摘要 第18届金砖国家峰会正在印度新德里举行，各国领导人陆续抵达Bharat Mandapam主会场，总理莫迪到场迎接与会者。来源：彭博社。
Narendra Modi arrives at the main convention hall as the summit gets underway (Source: Bloomberg)
中文摘要 印度总理纳伦德拉·莫迪抵达新德里Bharat Mandapam主会场，标志第18届金砖峰会正式启动。来源：彭博社。
HSBC Holdings Plc is set to start its search for a new chief financial officer, parsing through a pool of internal and external candidates willing and qualified to take on one of the biggest jobs in global finance. The bank is looking to replace Pam Kaur, who unexpectedly announced Thursday she woul
中文摘要 汇丰控股拟启动新任首席财务官搜寻，在内部和外部候选人中挑选接替Pam Kaur者，后者意外宣布离任。CFO为全球金融界最高职位之一。来源：彭博社。
The Economist invites applications for the Marjorie Deane internship
中文摘要 《经济学人》面向商业、金融与经济学方向开放Marjorie Deane实习计划申请，邀请相关领域新人提交申请。
Also in today’s newsletter: Hyrox gets a power boost
中文摘要 《金融时报》报道LIV高尔夫前途黯淡，相关问题转向法庭解决；同刊通讯另提及Hyrox赛事获得动力提升。
中文摘要 彭博社播客邀请Robert Friedland讨论全球铜资源严重短缺问题，探讨供需格局与影响。
Chinese President Xi Jinping arrives in Delhi for BRICS Summit 2026 ahead of bilateral talks (Source: Bloomberg)
中文摘要 中国国家主席习近平抵达印度新德里，出席2026年金砖国家峰会，并参加双边会谈。来源：彭博社。
India’s markets regulator proposed sweeping changes to the closing auction system, including a possible return to the previous method for settling derivatives on expiry days, after sharp price swings and liquidity concerns emerged around the month-old mechanism.
中文摘要 印度证券交易委员会（SEBI）针对上市一个月即引发剧烈波动和流动性问题的收盘竞价机制提出重大改革，或恢复到期日衍生品旧结算方式。来源：彭博社。
Belron is the sort of business that might help investors warm up to more deals
中文摘要 《金融时报》评论Belron（风挡玻璃巨头）潜在大型交易，认为此类业务有望修复欧洲低迷的IPO窗口，提振投资者信心。
Austrian central bank governor Martin Kocher says risk of elevated inflation is higher than it was a few months ago
中文摘要 奥地利央行行长Martin Kocher警告，高油价可能迫使欧洲央行进一步加息，当前通胀风险高于数月前的水平。来源：《金融时报》。
Seoul broadens its espionage legislation for the first time in more than seven decades
中文摘要 韩国首次自70多年来扩展其间谍法律，以保护半导体产业核心机密免受外国间谍活动威胁。来源：《金融时报》。
The wealth manager will scrutinise growth plans and aim to boost adviser morale
中文摘要 财富管理公司St James's Place拟审查增长计划，并着手提振近期人才流失后顾问团队的士气。来源：《金融时报》。
13 回复 · 程序员 节点
5 回复 · Apple 节点
11 回复 · 程序员 节点
24 回复 · 程序员 节点
23 回复 · 程序员 节点
48 回复 · 程序员 节点
6 回复 · Apple 节点
24 回复 · 程序员 节点
64 回复 · 程序员 节点
17 回复 · Apple 节点
该源今日无内容。