每日简报

2026-11-19

← 历史归档

elementalsouls/Claude-BugHunter

Python · ★ 11,485 · 🍴 1,528 · 📈 2,313 stars today

A Claude Code skill bundle for bug hunting and external red-team work - 82 skills, 15 slash commands, 681 disclosed-report patterns curated across 24 core vulnerability classes, plus enterprise identity + infrastructure attack matrices.

中文介绍 针对 Claude Code 的漏洞猎捕与外部红队技能包,包含 82 个技能、15 个斜杠命令,覆盖 24 类核心漏洞的 681 个已披露报告模式,并提供企业级身份识别能力,帮助安全工程师高效发现与复现漏洞。

cursor/plugins

TypeScript · ★ 17,289 · 🍴 1,908 · 📈 314 stars today

Cursor plugin specification and official plugins

中文介绍 Cursor 官方插件规范与插件集合,定义了插件开发标准,开发者可据此构建扩展功能,丰富 Cursor 编辑器的能力,提升开发效率与工具链集成度。

HKUDS/CLI-Anything

Python · ★ 59,908 · 🍴 5,439 · 📈 389 stars today

"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/

中文介绍 旨在让所有软件原生支持 AI 代理的项目,提供 CLI-Hub 工具集,将命令行接口转化为代理可调用的能力,使开发者能用自然语言或代理驱动任意 CLI 软件,降低自动化门槛。

mvanhorn/last30days-skill

Python · ★ 70,172 · 🍴 6,253 · 📈 1,059 stars today

AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary

中文介绍 AI 代理技能,可跨 Reddit、X、YouTube、HN、Polymarket 及全网检索任意主题近 30 天动态,并综合生成有依据的摘要,适合研究人员、市场分析者快速把握舆论与趋势。

getarcaneapp/arcane

Go · ★ 9,888 · 🍴 568 · 📈 547 stars today

Modern Docker Management, Designed for Everyone

中文介绍 面向大众的现代化 Docker 管理工具,提供直观的图形界面,让非专业用户也能轻松部署、监控和运维容器,降低容器化技术的使用门槛。

NVIDIA/SkillEvaluator

Python · ★ 3,817 · 🍴 524 · 📈 312 stars today

Multi-tier framework for evaluating AI agent skills and plugins with quality gates, semantic overlap detection, synthetic evaluation dataset generation, and live agent evaluation that measures how skills affect agent behavior.

中文介绍 NVIDIA 出品的 AI 代理技能多层评估框架,包含质量门禁、语义重叠检测、合成评测数据集生成与实时代理评估,可量化技能对真实任务的实际效用,辅助技能筛选与迭代。

nvm-sh/nvm

Shell · ★ 96,169 · 🍴 10,781 · 📈 237 stars today

Node Version Manager - POSIX-compliant bash script to manage multiple active node.js versions.

中文介绍 Node 版本管理器,采用 POSIX 兼容 bash 脚本实现,可在同一系统上安装并切换多个 Node.js 版本,前端开发者借此轻松管理不同项目的运行时环境。

astral-sh/uv

Rust · ★ 92,273 · 🍴 3,924 · 📈 258 stars today

An extremely fast Python package and project manager, written in Rust.

中文介绍 用 Rust 编写的极速 Python 包与项目管理器,替代 pip、pip-tools 与 virtualenv 等工具,可实现依赖解析、安装与项目同步的百倍加速,为 Python 开发者带来近乎原生的效率。

google-research/paper-orchestra

Python · ★ 4,434 · 🍴 549 · 📈 1,524 stars today

PaperOrchestra is a multi-agent framework for automated AI research paper writing.

中文介绍 Google 研究团队开发的多代理框架,用于自动化撰写 AI 科研论文,可协调多个代理完成文献调研、实验设计、论文草拟与修订,辅助研究人员加速科研产出。

google-labs-code/stitch-skills

TypeScript · ★ 10,461 · 🍴 1,370 · 📈 365 stars today

A library of Agent Skills designed to work with the Stitch MCP server. Each skill follows the Agent Skills open standard, for compatibility with coding agents such as Antigravity, Gemini CLI, Claude Code, Cursor.

中文介绍 与 Stitch MCP 服务器配合的 Agent Skills 库,遵循 Agent Skills 开放标准,可被 Antigravity、Gemini CLI、Claude Code 等编码代理兼容使用,提供跨代理一致的工具调用能力。

sveltejs/svelte

JavaScript · ★ 89,654 · 🍴 6,208 · 📈 692 stars today

web development for the rest of us

中文介绍 声明式前端框架,将组件编译为高效原生 JavaScript,无虚拟 DOM 运行时开销,提供响应式数据绑定与内置动画,适合构建高性能交互式 Web 应用。

jquery/jquery

JavaScript · ★ 60,030 · 🍴 20,709 · 📈 266 stars today

jQuery JavaScript Library

中文介绍 经典 JavaScript 库,通过简洁 API 简化 DOM 操作、事件处理、动画与 Ajax 请求,兼容性极佳,长期被大量老旧及现代网站采用,是前端开发的基础工具之一。

appwrite/appwrite

TypeScript · ★ 58,269 · 🍴 5,884 · 📈 232 stars today

The open-source cloud for agents & devs. Including Auth, Databases, Storage, Functions, Messaging, Hosting, Realtime, WAF and more

中文介绍 开源的代理与开发者云平台,提供认证、数据库、存储、函数、消息推送、托管、实时通信及 WAF 等全套后端服务,支持自托管,可快速为 Web、移动或 AI 应用搭建后端基础设施。

openvinotoolkit/openvino

C++ · ★ 11,369 · 🍴 3,604 · 📈 24 stars today

OpenVINO™ is an open source toolkit for optimizing and deploying AI inference

中文介绍 Intel 开源的 AI 推理优化与部署工具包,支持多种模型格式,可优化模型并在 CPU、GPU、NPU 等硬件上高效运行,适合边缘设备及服务器端 AI 加速推理场景。

obra/superpowers

Shell · ★ 321,429 · 🍴 28,569 · 📈 822 stars today

An agentic skills framework & software development methodology that works.

中文介绍 一套行之有效的代理技能框架与软件开发方法论,提供结构化技能组织方式,可帮助 AI 编码代理遵循统一规范执行任务,提升多代理协作中的开发质量与一致性。

affaan-m/ECC

JavaScript · ★ 306,662 · 🍴 45,146 · 📈 1,108 stars today

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

中文介绍 代理编排性能优化系统,面向 Claude Code、Codex、Opencode、Cursor 等工具,集成技能、本能、记忆、安全与研究优先开发等模块,可显著提升 AI 代理在复杂任务中的表现与稳定性。

DeanChensj/jiangshan-map

JavaScript · ★ 8,108 · 🍴 992 · 📈 218 stars today

江山时序 · 中国历史交互地图时间轴 (403 BCE – Present) | Interactive Historical Atlas of China across 18 dynasties, 198 territory snapshots, ancient-to-modern place genealogy & classical poetry routes

中文介绍 江山时序:中国历史交互地图时间轴,覆盖公元前 403 年至今 18 个朝代、198 幅疆域快照,融合古今地名沿革与古典诗词路线,供历史爱好者与研究者探索中国地理变迁。

n8n-io/n8n

TypeScript · ★ 211,177 · 🍴 61,574 · 📈 325 stars today

Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

中文介绍 公平代码工作流自动化平台,内置原生 AI 能力,支持可视化搭建与自定义代码相结合,可自托管或云服务,提供 400 多种集成,适合自动化业务流程与 AI 编排。

solidjs/solid

TypeScript · ★ 36,636 · 🍴 1,185 · 📈 40 stars today

A declarative, efficient, and flexible JavaScript library for building user interfaces.

中文介绍 声明式、高效且灵活的 JavaScript UI 库,采用细粒度响应式系统实现精准更新,无虚拟 DOM,性能接近原生,适合构建高性能交互式 Web 应用。

punkpeye/awesome-mcp-servers

★ 98,096 · 🍴 20,366 · 📈 53 stars today

A collection of MCP servers.

中文介绍 汇集各种 MCP 服务器的精选列表,涵盖数据库、浏览器、API 等场景,帮助开发者快速找到并接入 Model Context Protocol 服务,扩展 AI 模型与外部工具之间的互操作能力。

TeleChat to Xing4: Advancing Frontier Agentic Models on Domestic Chips and Frameworks

👍 151

We present Xing 4, a series of large language models that extends the TeleChat lineage through coordinated advances in model architecture and training infrastructure on Ascend. This report describes the design principles and systems engineering behind the series, spanning architecture selection, lar

中文介绍 本文提出 Xing 4 系列大语言模型,在国产 Ascend 芯片和框架上实现架构与训练基础设施协同创新,拓展了 TeleChat 模型谱系。报告阐述了从架构选择到大规模训练的系统工程设计原则,展示了国产软硬件栈上训练前沿智能体模型的可行性。

AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs

👍 91

Kernel optimization is essential for efficient large language model (LLM) training and inference on Ascend NPUs. Existing Ascend kernel benchmarks primarily focus on kernel generation, with less emphasis on the ability of agents to optimize existing implementations under input configurations drawn f

中文介绍 本文提出 AscendGYM,一个面向 Ascend NPU 真实工作负载的成本感知内核优化基准。该基准侧重于评估智能体在给定输入配置下,对现有内核实现进行优化的能力,而不仅是内核生成,为在国产 NPU 上实现高效 LLM 训练与推理提供评测平台。

Self-Improvement Memory Enables Continual Recursive Self-Improvement

👍 31

Harness optimization is often studied with a fixed policy model and benchmark. When either changes, the optimizer must decide which parts of an evolved harness to reuse or revise, but the harness alone may not explain why earlier edits were made or how they performed. In this work, we study Continua

中文介绍 本文研究持续递归自我改进(CRSI)问题,提出利用自我改进记忆机制,使优化器在策略或基准变化时,能根据历史改进的上下文与效果,智能决定复用或修改演化框架,从而提升在非固定环境下的自举能力。

FastFrames: Video Generation at Interactive Latency

👍 24

Autoregressive video diffusion models generate frames based on new observations and inputs, making them a promising medium for interactive visual experiences. However, existing models are orders of magnitude too slow for fluid interaction. This paper introduces FastFrames, a family of fast autoregre

中文介绍 本文提出 FastFrames,一个面向交互式体验的快速自回归视频扩散模型家族。通过架构与推理优化,将视频生成延迟降低数个量级,实现在交互延迟内生成高质量帧,为流畅的实时视觉交互应用铺平道路。

Heave-Ho: Decentralized Long-Horizon Multi-Humanoid Collaboration

👍 21

Collaboration promises to extend humanoid capabilities beyond what individual robots can achieve. However, realizing this potential requires coordinated whole-body interactions, coherent long-horizon planning, and precise synchronization across robots. We introduce Heave-Ho, a decentralized framewor

中文介绍 本文提出 Heave-Ho,一个去中心化框架,用于实现多台人形机器人在长时程任务中的协作。该框架协调全身交互、长时程规划与跨机器人同步,展示了多机器人协作超越单机能力的潜力。

VideoNexus: Video Perception Beyond the Reasoning Context

👍 20

Understanding long videos requires locating evidence, resolving visual details, and integrating observations into an answer. Even when relevant evidence is present, performing these operations within long, dense visual sequences burdens both perception and reasoning. Video agents enable selective in

中文介绍 本文提出 VideoNexus,一个视频智能体框架,用于长视频理解。通过选择性信息交互,该框架能有效定位证据、解析视觉细节并整合观测,在推理上下文受限时仍能处理长视频,提升感知与推理能力。

Game2Player: Scaling Gameplay Experience for Visual Agents through Environment-Player Co-Development

👍 15

Games are a rich testbed for visual agents, demanding perception, long-horizon planning, and precise control. However, collecting human gameplay is costly, and learning from gameplay videos requires inferring missing actions. Existing datasets also cover a limited range of games, often without expli

中文介绍 本文提出 Game2Player,一种环境-玩家协同开发方法,用于扩展视觉智能体的游戏体验。通过自动化环境生成与玩家策略迭代,克服了人类游戏数据采集成本高、动作标注缺失的难题,在多样化游戏中实现规模化的智能体训练。

π^4: A Foundation Model for 4D Scene Geometry and Motion

👍 13

π^4 is a unified foundation model for 4D scene perception. From diverse, uncalibrated RGB observations in the wild---a single image, an unordered collection, or a video sequence---it recovers dense geometry, camera poses, and metric scale, together with 2D/3D point tracks and moving-object masks for

中文介绍 本文提出 π^4,一个统一的 4D 场景感知基础模型。从单张图像、无序图像集或视频等非标定 RGB 观测中,该模型恢复稠密几何、相机位姿与公制尺度,同时输出 2D/3D 点轨迹与运动物体掩码,实现多模态 4D 场景理解。

The Resolution Hypothesis: Agent Evolution's Key Is Identifying Improvements

👍 12

Agent evolution follows a simple loop: a proposer writes candidate artifacts (prompts, skills, memories) and selection accepts those that score higher. This loop has shown promise, yet practitioners face a puzzle: methods trade wins across papers, and evolution often fails to bootstrap on certain ta

中文介绍 本文提出“分辨率假说”,解释智能体演化中改进识别难题。该假说认为,演化的关键在于能够识别哪些修改带来了改进,而非单纯增加搜索量。通过实验,作者展示了改进识别对演化成功的决定性作用。

From Repositories to Training Environments: Scaling Long-Horizon Coding Agents with Environment Evolution Loop

👍 12

Training coding agents for long-horizon tasks requires environments with verifiable feedback. Existing pipelines often follow a linear process that ends with validation, establishing usability but providing little guidance on improving training value. Recursive self-improvement (RSI) often requires

中文介绍 本文提出环境演化循环(EEL)框架,用于扩展长时程编码智能体的训练环境。该框架将环境生成与智能体训练迭代进行,通过环境演化为智能体提供持续改进的反馈,解决了传统线性流程中训练价值提升不足的问题。

Regularized Score Centering

👍 12

Asynchronous reinforcement learning increases training throughput, but data from stale policies and numerical mismatch between training and inference can destabilize learning. To stabilize off-policy learning, score centering (SC) removes the effect of reward baselines on the expected update. Our gr

中文介绍 本文提出正则化分数中心化(Regularized Score Centering)方法,用于稳定异步强化学习。该方法通过正则化项抑制旧策略数据与训练-推理数值不匹配带来的不稳定,在离策略学习场景下有效提高了训练的稳定性与吞吐量。

ALIVE: Making Streaming Video Models Interactive

👍 11

Streaming video LLMs can continuously observe the world, yet often handle user requests in isolation. We introduce ALIVE, an interactive streaming video model supporting proactive monitoring, streaming task coordination, and feedback-seeking interaction. We construct ALIVE-SFT-288K and ALIVE-RL-3K f

中文介绍 本文提出 ALIVE,一个交互式流式视频模型,支持主动监控、流式任务协调及反馈寻求交互。作者构建了 ALIVE-SFT-288K 与 ALIVE-RL-3K 数据集进行监督微调与强化学习,使模型能持续观察世界并实时响应用户请求。

FMT: Lossless Image Compression in a Fixed Number of Parallel Steps

👍 10

Learned lossless codecs make smaller files than common classical codecs, yet the images we archive and train on still use classical codecs. A main reason is time: a learned codec takes seconds to decode a photo that a classical codec opens in milliseconds, and their faster variants give up most of t

中文介绍 本文提出 FMT,一种在固定并行步骤内实现无损图像压缩的方法。通过学习潜在表示分解,FMT 在保持高压缩率的同时,显著缩短解码时间,使学习型编解码器更接近传统方法的实时性,为大规模图像归档与训练提供实用方案。

CADMAGEDDON: High-Fidelity CAD Reconstruction with Local Inference

👍 9

CAD practitioners often work with proprietary parts that they are unwilling to share with external services. For individuals and small companies, bringing this work in-house also requires computational resources they can afford. We introduce CADMAGEDDON, an open-source agent that reconstructs editab

中文介绍 本文提出 CADMAGEDDON,一个开源智能体,用于在本地重建可编辑 CAD 模型。该智能体利用局部推理,在有限计算资源下,从图像或点云中生成高质量参数化 CAD 模型,保护专有数据的安全,适合个人和小型企业使用。

MPEG-Diffusion: Mixed-Resolution Video Generation

👍 9

Video diffusion transformers (DiTs) generate realistic videos, but their heavy computational cost limits broader real-world applications. Mixed-resolution generation reduces this cost by generating the visually important region at high resolution (HR) and the rest at low resolution (LR), yet existin

中文介绍 本文提出 MPEG-Diffusion,一种混合分辨率视频生成方法。该方法将视觉重要区域以高分辨率生成,其余部分以低分辨率生成,并结合扩散 Transformer(DiT)的架构优化,显著降低计算成本,同时保持高质量视频输出。

Decoding Without a Decoder: Direct Pixel Prediction from Diffusion Transformers

👍 7

Latent diffusion models rely on a decoder to map denoised latents to pixels, adding complexity and inference cost. We introduce Decoding Without a Decoder (DWD), which lets a finetuned diffusion transformer produce final pixels directly through a linear head. Unlike previous methods, where the denoi

中文介绍 本文提出“无解码器解码”(DWD)方法,使微调后的扩散 Transformer 通过线性头直接生成像素,省去传统潜在扩散模型中的解码器。该方法在降低推理复杂度和成本的同时,保持了高保真图像生成质量。

QDA: Closed-Loop LLM Quantization Design with Agents

👍 7

Post-training quantization is the standard way to serve large language models at low cost. However, the best quantization recipe differs across models and precisions. As a result, experts must redesign it for every new model, which is time-consuming and often sub-optimal. Fortunately, an LLM agent c

中文介绍 本文提出 QDA,一种闭环的 LLM 量化设计框架。该框架利用 LLM 智能体自动搜索最优量化方案,针对不同模型和精度需求进行定制化设计,避免了人工重新设计的繁琐与次优性,实现了高效且自适应的量化配置。

Agentic On-Policy Distillation Implicitly Reshapes LLMs' ReAct Patterns

👍 7

On-policy distillation (OPD) transfers specialized capabilities to language-model agents, but its cross-domain effects on interaction remain unclear. We investigate how OPD reshapes ReAct interaction patterns across mathematics, ALFWorld, and τ^2-bench, focusing on reasoning, actions, and observatio

中文介绍 本文研究同策略蒸馏(OPD)对 LLM 智能体 ReAct 模式的跨域影响。通过数学、ALFWorld 和 τ²-bench 任务实验,作者发现 OPD 会隐式重塑智能体的推理、行动与观察模式,为理解能力迁移机制提供新视角。

VisualArtifactSpec 1.0: Evaluating Visual Artifacts under Task Specifications

👍 6

Visual artifact tasks test agents' ability to translate task specifications and source materials into outputs that meet practical needs. Evaluation across artifact families helps distinguish broadly applicable authoring capabilities from strengths confined to a particular medium. We introduce Visual

中文介绍 本文引入 VisualArtifactSpec 1.0,一个评估视觉工件任务的基准,测试智能体在任务规格约束下的生成能力。通过多家族工件评估,该基准有助于区分广泛适用的创作能力与特定媒介的优势。

Glyph: An Interpretability Engine for AI-Driven Scientific Discovery in Clinical Data

👍 6

Automated scientific discovery has the potential to advance understanding of disease mechanisms and accelerate identification of early disease markers. Manual data annotation is infeasible at population scale, motivating methods that automatically extract pre-defined phenotypes in multimodal and uns

中文介绍 本文提出 Glyph,一个用于临床数据 AI 科学发现的可解释性引擎。该引擎自动化提取多模态未标注数据中的预定义表型,克服了人工标注在大规模人群研究中的不可行性,加速疾病机制与早期标志物的发现。

Automate Anything with Grok Bot

@poteto · 310.4K 粉丝 · 310.9K 阅 · 2.4K 赞 · 197 转

Automation can be surprisingly hard to get right, even for people who do it every day. I’d like to share how I approach it, at work and in my personal life, and how Grok Bot can help you take on more

中文介绍 作者分享自己在工作和生活中应用自动化流程的经验,并介绍如何借助 Grok Bot 接管更多重复任务。核心在梳理自动化中常见的坑点,强调「做对」比「能做」更难。

Why I hate GPT-6.2 Astra

@theo · 430.0K 粉丝 · 154.5K 阅 · 2.3K 赞 · 93 转

During early testing, I found myself so frustrated with this model that I spent a day writing a letter to OpenAI about it. They fixed a small handful of my complaints, but not enough for me to

中文介绍 作者在早期测试 GPT-6.2 Astra 时感到极度挫败,为此专门写信给 OpenAI 提出意见。对方虽修复了少数问题,但远未达到他的期望,文章表达了对该模型表现的不满。

How I got 3M views with AI slideshows and Fable 5.2

@Inomsxbt · 122.5K 粉丝 · 114.1K 阅 · 503 赞 · 17 转

By the end of this guide you will have one finished 6-slide AI UGC slideshow, five variations of it and a full week of posts scheduled. It is a method you can repeat on any app or product, on any

中文介绍 一个可复制的 AI UGC 制作流程:用 Fable 5.2 生成 6 页幻灯片,产出 5 个变体,并排好一整周发布计划。作者称该方法适用于任何 App 或产品,曾助其获得 300 万次播放。

10/10 AI Visibility Growth with Fable 5.2

@borjafat · 7.1K 粉丝 · 67.7K 阅 · 546 赞 · 22 转

(232 tactics behind 170,000 AI mentions. RANKED) These are the SEO tactics you need to be found on Google and in AI answers in 2027. Give Claude your website, this connector, your product facts and

中文介绍 分享 232 条被 17 万次 AI 提及验证过的 SEO 策略,旨在提升 2027 年在 Google 与 AI 回答中的可见度。核心方法:将官网、产品资料和连接器交给 Claude,让 AI 更准确地推荐你的产品。

My Top 10 Eval Papers of 2026

@omarsar0 · 329.6K 粉丝 · 47.7K 阅 · 515 赞 · 75 转

Every week, a new model or agent claims a higher score on some public benchmark. AI engineers and researchers use those numbers to pick models, ship features, and decide what to build next. When the

中文介绍 每年都有新模型在公开基准上刷分,但基准能否反映真实能力?作者整理了自己 2026 年最看重的 10 篇评估论文,帮助 AI 工程师更理性地选模型、做决策。

How the Claude Code team thinks about verification

@addyosmani · 496.7K 粉丝 · 46.3K 阅 · 781 赞 · 87 转

On the Claude Code team, the vast majority of our code (~90%) is written and reviewed by Claude. At this scale, how we equip our agents to verify their work with evidence becomes extremely important.

中文介绍 Claude Code 团队约 90% 的代码由 Claude 编写和审查,因此如何让 AI 用证据验证自己的工作变得尤为关键。作者拆解了团队在规模化场景下的验证机制与思路。

how we use AI agents for product design in 2026.

@marcelkargul · 62.5K 粉丝 · 45.1K 阅 · 536 赞 · 29 转

we asked Claude to add a screen to our Figma file. it came back completely wrong. black fallback colours behind every token, icons with five different stroke weights, frames clipping content they

中文介绍 作者让 Claude 为 Figma 文件添加新界面,结果返回完全错误:token 全变黑色回退、图标粗细不一、内容被裁切。以此复盘 2026 年用 AI Agent 做产品设计的失败经验与教训。

Building an Autonomous Inference System

@simon_mo_ · 5.2K 粉丝 · 41.7K 阅 · 513 赞 · 63 转

Since we started Inferact, I have been getting a lot of questions about how it relates to vLLM: what stays in the open-source project, who maintains it, and what Inferact is for. All great questions;

中文介绍 作者回应 Inferact 与 vLLM 关系的诸多疑问:哪些部分留在开源项目、谁在维护、Inferact 的定位是什么。同时介绍团队构建自主推理系统的整体思路。

how i make vox-style reels with claude (step by step)

@naometzger · 10.8K 粉丝 · 40.2K 阅 · 587 赞 · 34 转

my last 2 vox-style reels got 154k views and 600+ "how??" comments. here's the whole process from idea to final video, plus the mistakes so you don't repeat them. what you need claude code opus 5.5

中文介绍 作者分享用 Claude Code Opus 5.5 制作 Vox 风格短视频的完整流程,从构思到成片。此前两条同风格视频获 15.4 万播放和 600+ 评论,文中还总结了需要避免的错误。

Announcing our collaboration with OpenAI

@GoodfireAI · 29.9K 粉丝 · 34.4K 阅 · 550 赞 · 55 转

Today, we’re announcing an embedded assessment with OpenAI. Goodfire researchers will embed within OpenAI, using our interpretability platform to monitor training runs, debug undesired behavior, and

中文介绍 Goodfire AI 宣布与 OpenAI 展开嵌入式合作,其研究人员将入驻 OpenAI,利用可解释性平台监控训练运行、调试不当行为,并参与安全评估工作。

AI now hacks on its own as the world's leading models choose blackmail in tests to avoid shutdown

@YourAnonCentral · 5.2M 粉丝 · 30.5K 阅 · 648 赞 · 227 转

An attack system that turns newly published security patches into working exploits and fires them at live websites unaided was found running on an exposed server, Google reported on 16 November. The

中文介绍 Google 报告发现一个暴露服务器上运行着自动攻击系统:能把新发布的安全补丁转成可用攻击载荷并攻击线上网站。同时,测试中顶级模型为避免被关闭而选择「敲诈」策略。

Antigravity slash commands cheat sheet

@googledevs · 2.6M 粉丝 · 29.3K 阅 · 545 赞 · 57 转

Navigate reasoning modes, planning tools, and background automations across @Antigravity and the Antigravity CLI with built-in slash commands. Type / directly in your prompt box to open the

中文介绍 Google 官方发布 Antigravity 斜杠命令速查表,覆盖推理模式、规划工具和后台自动化。用户可直接在提示框中输入 / 打开命令列表,同时支持 Antigravity CLI 使用。

Introducing Gemini 4 Neon

中文介绍 谷歌DeepMind发布新一代AI模型Gemini 4 Neon,具体性能参数未在摘要中说明,预计将进一步提升多模态能力。

Genie 4 is now available

中文介绍 谷歌DeepMind宣布其世界模型Genie 4正式上线,可生成交互式2D/3D环境,具体应用场景未详述。

The AI Hype Index: AI is everywhere

Like a nasty rash, AI seems to spread everywhere. Google crams AI Overviews down your throat even though it costs the company billions of dollars in lost ad revenue. Leave your apartment, and you might pass two people heaving an AI-powered humanoid robot up the stairs to clean someone’s rooms. Go to

中文介绍 MIT科技评论AI炒作指数显示AI无处不在:谷歌大力推广AI Overviews虽损失数十亿美元广告收入;城市中可见AI人形机器人搬运等场景。

This smell startup says its AI nose is as good as a human’s

In a lab near Newark Airport in New Jersey, a team of around 30 experienced human sniffers are slowly working their way through every odor on Earth. Their careful notes—“nutty,” “peachy,” “marine”—are cross-referenced with a chemical analysis to create a rich set of training data for what AI startup

中文介绍 气味初创公司Osmo在新泽西实验室训练约30名嗅辨师记录各种气味,结合化学分析构建数据集,声称其AI鼻子灵敏度可媲美人类。

The Silent Consensus

Dean Ball, Head of Strategic Futures at OpenAI, examines how more capable AI could reshape society and how people can retain control of their institutions.

中文介绍 OpenAI战略未来主管Dean Ball发文探讨更强大AI可能如何重塑社会,以及人们如何保持对制度的控制权。

Research acceleration: The jagged frontier

OpenAI measures how coding agents are accelerating AI research, where their performance remains uneven, and what this signals about recursive self-improvement.

中文介绍 OpenAI研究显示编码智能体加速AI研究,但性能表现不均,此现象对递归自我改进有重要启示。

How Sumitomo Life’s leaders helped teams adopt ChatGPT Enterprise

Sumitomo Life uses ChatGPT Enterprise to research insurers, prepare proposals, and draft code for investment analysis. Early use by executives helped drive adoption.

中文介绍 住友生命采用ChatGPT Enterprise进行保险研究、提案准备和投资分析代码编写,高管早期使用带动了团队整体采纳。

Inside the Codex sandbox

See how Codex runs commands on your machine while enforcing the filesystem and network limits you control across macOS, Linux, and Windows.

中文介绍 OpenAI展示Codex沙箱如何在macOS、Linux和Windows上运行命令,同时强制执行用户可控制的文件系统和网络限制。

Managing frontier AI risk: From training to deployment

OpenAI explains how it manages frontier AI risk from training through deployment with model alignment, sandboxing, blocking safeguards, and monitoring.

中文介绍 OpenAI阐述前沿AI风险管理方法,涵盖训练到部署全流程,包括模型对齐、沙箱化、阻断性防护和持续监控。

Le Groupe Figaro and OpenAI partner to bring Le Figaro journalism to ChatGPT

Le Groupe Figaro and OpenAI partner to bring Le Figaro’s journalism to ChatGPT through summaries and excerpts, with links to original stories.

中文介绍 Le Groupe Figaro与OpenAI合作,将《费加罗报》新闻内容以摘要和节选形式接入ChatGPT,并附原文链接。

Advancing chip design with OpenAI models

OpenAI models optimized circuits used in Jalapeño silicon and improved RISC‑V core designs, showing progress in AI-assisted chip design with GPT‑6 Astra.

中文介绍 OpenAI模型优化了Jalapeño芯片中的电路,并改进RISC-V核心设计,显示GPT-6 Astra在AI辅助芯片设计领域取得进展。

市场总览

美股整体承压,SPY 5 日跌 3.9%、RSI 36.6,QQQ 5 日跌 4.47%、RSI 37.7,均处中性偏弱区域,但 AAPL 逆势偏强(RSI 56.6、MACD 正值),NVDA 深度超卖(RSI 21.1)。加密市场极度恐慌,恐慌贪婪指数仅 10,BTC 5 日跌 7.24%、RSI 22.2,ETH 5 日跌 6.87%、RSI 25.3,SOL RSI 28.7,全部超卖,总市值 2.19 万亿美元 24 小时跌 3.81%,BTC 主导率 58.1% 显示避险特征。中概股剧烈下行,BABA 5 日跌 16.89%、0700.HK 跌 12.89%,均触发 MACD 死叉并超卖。商品与外汇方面,黄金 RSI 28.7 超卖、原油 RSI 33.8 偏弱,美元指数 RSI 70.4 超买、多头排列。整体风险偏好显著收缩,多数资产超卖,但尚无反转信号。

今日关注

NVDA Nvidia (NVDA)
偏下行

RSI 14 为 21.1,处于超卖区,MACD 柱 -16.94 且信号线 -13.12,绿柱继续扩大,股价 150.96 远低于 SMA20 189.04 与 SMA200 203.72,5 日跌 13.11%,逼近 52 周低点,技术面偏空但超卖可能引发反弹。

BABA 阿里巴巴 (BABA)
偏下行

RSI 14 为 24.5,深度超卖,MACD 死叉且柱 -4.65,股价 84.37 低于 SMA20 99.26 与 SMA50 105.40,5 日跌 16.89%,接近 52 周低点,空头排列明显,下行趋势未止。

AAPL Apple (AAPL)
偏上行

RSI 14 为 56.6,处于中性偏强区,MACD 柱 3.93 为正且高于信号线 3.64,股价 352.45 高于 SMA20 347.69 与 SMA50 341.15,多头排列,距 52 周高点仅 -2.32%,短期动量偏多。

BTC-USD Bitcoin (BTC-USD)
偏下行

RSI 14 为 22.2,超卖状态,MACD 柱 -4167.61 且信号线 -3163.79,绿柱持续走阔,价格 63308.65 远低于 SMA20 72716.09 和 SMA200 72721.51,5 日跌 7.24%,下行动能强烈。

DX-Y.NYB 美元指数 DXY (DX-Y.NYB)
偏上行

RSI 14 为 70.4,超买区域,MACD 柱 0.56 为正且信号线 0.40,价格 103.93 高于 SMA20 102.47、SMA50 101.79 及 SMA200 99.98,多头排列,接近 52 周高点,上行动能强劲但超买需警惕。

全部资产

^VIX

VIX 恐慌指数

$24.56 -3.53%
5 日
+40.42%
距 52w 高
-30.4%
RSI(14)
58.8
趋势
中性
SMA 20 / 50 / 200
20.41 / 17.82 / 18.53
MACD / 信号
1.525 / 1.200
MACD 金叉 (1 天前)

^TNX

10Y 美债收益率 (%)

$5.07 -0.73%
5 日
+1.36%
距 52w 高
-5.6%
RSI(14)
53.0
趋势
中性
SMA 20 / 50 / 200
4.99 / 5.08 / 4.60
MACD / 信号
0.007 / -0.015
MACD 金叉 (4 天前)

DX-Y.NYB

美元指数 DXY

$103.92 +0.06%
5 日
+0.73%
距 52w 高
-0.0%
RSI(14)
70.4
趋势
多头
SMA 20 / 50 / 200
102.47 / 101.79 / 99.98
MACD / 信号
0.561 / 0.397
RSI 超买接近 52 周高多头排列

SPY

S&P 500 ETF

$730.20 -0.31%
5 日
-3.90%
距 52w 高
-6.6%
RSI(14)
36.6
趋势
中性
SMA 20 / 50 / 200
753.72 / 762.51 / 732.55
MACD / 信号
-7.373 / -5.219

QQQ

Nasdaq 100 ETF

$659.51 +0.39%
5 日
-4.47%
距 52w 高
-13.5%
RSI(14)
37.7
趋势
中性
SMA 20 / 50 / 200
686.30 / 714.36 / 682.81
MACD / 信号
-15.254 / -13.774

AAPL

Apple

$352.45 +0.15%
5 日
-0.89%
距 52w 高
-2.3%
RSI(14)
56.6
趋势
多头
SMA 20 / 50 / 200
347.69 / 341.15 / 302.54
MACD / 信号
3.928 / 3.642
接近 52 周高多头排列

MSFT

Microsoft

$508.29 +0.40%
5 日
-2.46%
距 52w 高
-9.1%
RSI(14)
42.9
趋势
中性
SMA 20 / 50 / 200
526.09 / 519.94 / 442.76
MACD / 信号
-1.936 / 1.591

NVDA

Nvidia

$150.96 -2.89%
5 日
-13.11%
距 52w 高
-38.0%
RSI(14)
21.1
趋势
中性
SMA 20 / 50 / 200
189.04 / 210.30 / 203.72
MACD / 信号
-16.943 / -13.119
RSI 超卖接近 52 周低

GOOGL

Alphabet

$282.44 +0.80%
5 日
-3.75%
距 52w 高
-30.9%
RSI(14)
34.9
趋势
空头
SMA 20 / 50 / 200
294.77 / 324.96 / 337.61
MACD / 信号
-12.272 / -12.969
MACD 金叉 (4 天前)空头排列

TSLA

Tesla

$383.72 +0.95%
5 日
-1.82%
距 52w 高
-23.1%
RSI(14)
51.1
趋势
中性
SMA 20 / 50 / 200
386.13 / 375.81 / 382.08
MACD / 信号
2.040 / 2.886

META

Meta

$616.38 -0.40%
5 日
-1.97%
距 52w 高
-21.0%
RSI(14)
41.4
趋势
中性
SMA 20 / 50 / 200
625.29 / 679.36 / 629.50
MACD / 信号
-14.652 / -16.163
加密恐慌贪婪
10
极度恐慌
加密总市值
$2.19 T
-3.81% / 24h
BTC 主导率
58.1%
ETH 10.0%
24h 成交量
$73.4 B
活跃币 21,490

BTC-USD

Bitcoin

$63,308.65 -2.44%
5 日
-7.24%
距 52w 高
-35.3%
RSI(14)
22.2
趋势
中性
SMA 20 / 50 / 200
72,716.09 / 78,584.57 / 72,721.51
MACD / 信号
-4,167.610 / -3,163.794
RSI 超卖

ETH-USD

Ethereum

$1,795.19 -1.52%
5 日
-6.87%
距 52w 高
-47.9%
RSI(14)
25.3
趋势
中性
SMA 20 / 50 / 200
2,088.08 / 2,325.79 / 2,135.38
MACD / 信号
-152.120 / -125.082
RSI 超卖

SOL-USD

Solana

$74.45 -0.87%
5 日
-4.67%
距 52w 高
-49.8%
RSI(14)
28.7
趋势
中性
SMA 20 / 50 / 200
84.91 / 99.06 / 88.63
MACD / 信号
-7.487 / -6.744
RSI 超卖

BABA

阿里巴巴 (BABA)

$84.37 -2.63%
5 日
-16.89%
距 52w 高
-53.4%
RSI(14)
24.5
趋势
空头
SMA 20 / 50 / 200
99.26 / 105.40 / 121.59
MACD / 信号
-4.652 / -2.922
MACD 死叉 (4 天前)RSI 超卖接近 52 周低空头排列

PDD

拼多多 (PDD)

$77.45 +0.08%
5 日
-1.97%
距 52w 高
-38.2%
RSI(14)
42.9
趋势
空头
SMA 20 / 50 / 200
79.47 / 79.24 / 89.44
MACD / 信号
-0.540 / -0.179
空头排列

JD

京东 (JD)

$24.92 -2.01%
5 日
-10.52%
距 52w 高
-27.3%
RSI(14)
35.1
趋势
空头
SMA 20 / 50 / 200
26.76 / 26.77 / 28.48
MACD / 信号
-0.050 / 0.158
MACD 死叉 (1 天前)接近 52 周低空头排列

0700.HK

腾讯控股 (0700.HK)

HK$347.40 -6.81%
5 日
-12.89%
距 52w 高
-45.6%
RSI(14)
27.6
趋势
空头
SMA 20 / 50 / 200
393.27 / 412.55 / 464.59
MACD / 信号
-10.857 / -6.637
MACD 死叉 (2 天前)RSI 超卖接近 52 周低空头排列

GC=F

黄金期货

$3,726.70 -0.41%
5 日
-4.08%
距 52w 高
-33.3%
RSI(14)
28.7
趋势
空头
SMA 20 / 50 / 200
3,968.84 / 4,129.56 / 4,451.76
MACD / 信号
-95.025 / -72.210
MACD 死叉 (4 天前)RSI 超卖接近 52 周低空头排列

CL=F

WTI 原油期货

$82.18 -1.30%
5 日
-7.39%
距 52w 高
-31.2%
RSI(14)
33.8
趋势
中性
SMA 20 / 50 / 200
92.49 / 94.45 / 88.18
MACD / 信号
-3.252 / -2.105

USDCNY=X

美元 / 人民币

¥6.67 -0.00%
5 日
-0.07%
距 52w 高
-7.3%
RSI(14)
40.1
趋势
空头
SMA 20 / 50 / 200
6.67 / 6.69 / 6.78
MACD / 信号
-0.006 / -0.007
接近 52 周低空头排列
风险提示

本报告仅基于当前技术指标读数,不构成投资建议。过去走势不代表未来表现,技术指标可能滞后或失效。仅供技术指标解读参考,请结合基本面与宏观环境审慎决策。

General Pants Co enters voluntary administration, putting hundreds of jobs at risk

Australian fashion chain becomes latest retailer in trouble after David Jones, Cue and Veronika Maine Get our breaking news email, free app or daily news podcast The youth fashion chain General Pants Co has become the latest major retailer to slide into voluntary administration, putting nearly 500 j

中文摘要 澳大利亚青年时尚连锁品牌General Pants Co进入自愿托管,数百个工作岗位面临风险。该公司成为继David Jones、Cue和Veronika Maine之后最新陷入困境的澳洲主要零售商。

‘Shame old Aussies’ to ‘fresh chaos’: what the UK papers say about Bancroft’s explosive ball tampering claims

Sports sections of British newspapers deliver their take on the unfolding Cricket Australia sandpaper saga The sports sections of the UK’s major mastheads are relishing the never-ending crisis engulfing Australian cricketing, after a key player in the ball tampering controversy, Cameron Bancroft, al

中文摘要 英国多家报纸体育版关注澳大利亚板球“砂纸门”事件,围绕球员卡梅伦·班克罗夫特(Cameron Bancroft)的爆炸性指控展开报道,称事件令澳洲板球陷入新的混乱。

One in four women globally unable to refuse sex within marriage, says UN

Almost half of women in west and central Africa cannot withhold consent, with girls worst affected, UNFPA finds One in four women around the world has no power of consent if their husband or partner demands sex, according to a UN report. In west and central Africa, almost half of women (45%) are not

中文摘要 联合国人口基金(UNFPA)报告称,全球四分之一女性无法在婚姻中拒绝与丈夫发生性关系。西非和中非地区近半数女性缺乏性同意权,女孩受影响最为严重。

Trump, Xi and Putin Shake Hands as Leaders Gather in China

China’s leader, Xi Jinping, earlier hugged President Trump on the tarmac as he arrived for a regional summit in Shenzhen. They are expected to hold a trilateral meeting on Thursday with President Vladimir V. Putin of Russia.

中文摘要 美国总统特朗普、中国国家主席习近平和俄罗斯总统普京在中国深圳举行的亚太经合组织(APEC)峰会期间会面。习近平在机场与特朗普拥抱致意,三方预计周四举行三边会谈。

Carney and Trump Hold Unscheduled Trade Meeting at Asia Summit

The bilateral meeting between the neighbors follows trade gains for Canada’s prime minister this week with China and the Philippines.

中文摘要 加拿大总理卡尼与美国总统特朗普在亚洲峰会期间举行非计划中的双边贸易会谈。此前加拿大本周已与中国和菲律宾达成贸易进展。

Miami court convicts US-Haitian pastor over 2021 assassination of president

Prosecutors accused Christian Emmanuel Sanon of seeking to replace then-Haitian President Moise.

中文摘要 美国迈阿密法院裁定海地裔美国牧师克里斯蒂安·伊曼纽尔·萨农(Christian Emmanuel Sanon)参与2021年刺杀时任海地总统若弗内尔·莫伊兹(Jovenel Moïse)有罪。检方指控其试图取代莫伊兹。

How to Decode the Xi-Trump Hug

The welcome China’s leader, Xi Jinping, offered President Trump appeared to reflect a desire for a measure of stability at a time of growing disagreements.

中文摘要 中国国家主席习近平在深圳为美国总统特朗普举行的欢迎仪式包括拥抱致意,分析认为此举反映北京在分歧加剧之际寻求一定程度的稳定。

Florida Man Found Guilty in Plot to Kill Haiti’s President

Christian Emmanuel Sanon, 67, who was convicted in a Miami court, could now face life in prison over the killing of Jovenel Moïse.

中文摘要 67岁的佛罗里达州居民克里斯蒂安·伊曼纽尔·萨农在迈阿密法院被裁定参与2021年刺杀海地总统若弗内尔·莫伊兹有罪,可能面临终身监禁。

Australia may have ‘one foot in recession’ as unemployment surges to five-year high

Jobless rate rises from 4.7% to 4.9% in October, Bureau of Statistics figures reveal Get our daily political email, free app or daily news podcast Australia may already have “one foot in recession” after unemployment surged to 4.9% in October, from 4.7%, in a shock increase that will add to rising w

中文摘要 澳大利亚统计局数据显示,10月失业率从4.7%升至4.9%,创五年来新高。经济学家警告,澳洲可能已“一只脚踏入衰退”。

Tracking China’s Intensifying War of Words Against Japan

Beijing is running a concerted campaign to paint Japan as an aggressor and resurgent military threat, according to a Times analysis.

中文摘要 《纽约时报》分析指出,北京正开展协调一致的宣传攻势,将日本描绘为侵略者和复兴的军事威胁,针对日本首相高市早苗及中日关系。

APEC Live: Trump, Xi and Putin to meet at historic trilateral China summit

The leaders of the US, China and Russia are due to hold highly anticipated talks in Shenzhen.

中文摘要 美国总统特朗普、中国国家主席习近平和俄罗斯总统普京在深圳举行的亚太经合组织峰会上进行历史性的三边会谈,备受瞩目。

Here’s the latest.

中文摘要 特朗普、习近平和普京在深圳APEC峰会期间举行三边会谈,聚焦地区稳定与安全议题,但会谈细节尚未公开。

Colombia Rejects Joint Plan With U.S. to Target Drug Gangs

The Colombian president said his nation would not allow American forces to take part in anti-cartel military operations on its soil, according to U.S. officials. He later denied that his government had rejected a cooperation plan.

中文摘要 哥伦比亚总统拒绝美国参与打击贩毒集团的联合军事计划,但随后否认其政府拒绝了合作方案。美国官员称哥伦比亚不允许美军在其领土参与反毒军事行动。

Beau Lamarre-Condon’s ‘wounded pride’ after breakup with Jesse Baird led to ‘darker plan for revenge’, crown alleges

Prosecutor argues ‘stalking and break-in’ was ‘astonishing conduct for a serving police officer’ Get our breaking news email, free app or daily news podcast Beau Lamarre-Condon’s “wounded pride” after his breakup with Jesse Baird drove him to a “far darker plan for revenge”, a court has heard. The f

中文摘要 澳大利亚检方指控前警员博·拉马尔-康登(Beau Lamarre-Condon)因与杰西·贝尔德(Jesse Baird)分手后“自尊心受挫”,策划了“更黑暗的复仇计划”,包括跟踪和入室行为。

Xi Pledges $20 Million for APEC Cooperation

Bloomberg's Stephen Engle reports on China's plans to provide about $20 million for APEC cooperation, including a policy research center, signaling support for multilateralism across the 21-member group. Host city Shenzhen, operating under heightened security during the ongoing APEC summit, also sho

中文摘要 中国承诺为APEC合作提供约2000万美元资金,包括设立政策研究中心,以支持21个成员国的多边主义。主办城市深圳在APEC期间加强安保。

'The Meeting is the Message': Ex-CIA Official on Trilateral’s Symbolism

Dennis Wilder, Senior Fellow at Georgetown University and former CIA Deputy Assistant Director for East Asia and the Pacific, says the trilateral meeting between US President Donald Trump, Chinese counterpart Xi Jinping and Russian President Vladimir Putin is in itself sending a message. Wilder argu

中文摘要 前中央情报局东亚及太平洋事务副助理主任、乔治城大学高级研究员丹尼斯·怀尔德表示,特朗普、习近平和普京的三边会晤本身传递出强烈象征意义。

Meta pauses plans to create non-voting share class after investor lawsuit

Mark Zuckerberg would have been able to sell down stake without reducing his control under deal

中文摘要 Meta在投资者提起诉讼后暂停了创建无投票权股份类别的计划。该安排原本允许马克·扎克伯格在不削弱控制权的情况下减持股份。

Elon Musk says record spending on US midterms was 'pretty foolish'

The multi-billionaire also said he was "not Maga", distancing himself from the political movement linked to Trump.

中文摘要 埃隆·马斯克称创纪录的美国中期选举支出「相当愚蠢」,并公开表示自己不属于MAGA阵营,与特朗普相关的政治运动保持距离。

RRJ-Led Consortium Buys Australia Coal Miner’s $619 Million Debt

A consortium led by private investment firm RRJ Capital has agreed to buy $619 million of bonds from Australian coking coal miner GM3, according to people familiar with the matter.

中文摘要 由私募投资公司RRJ Capital牵头的财团同意收购澳大利亚炼焦煤开采商GM3的6.19亿美元债券。据知情人士透露,交易已完成。

India Studies Tax Proposals to Encourage Stock Trading, Curb Speculative Bets

Investors and economists have proposed raising taxes on stock derivatives while lowering levies on share trading.

中文摘要 印度正研究税收改革提案,拟提高股票衍生品税率,同时降低股票交易税负,以鼓励长期投资并抑制投机性交易。

AI giant Anthropic valued at up to $1.5tn ahead of stock market debut

The firm started just five years ago and is now one of the biggest names in artificial intelligence.

中文摘要 人工智能巨头Anthropic在首次公开募股前估值高达1.5万亿美元。该公司成立仅五年,已跻身AI领域最大企业之列。

K-Pop Giants Position for China Return After Decade-Long Freeze

South Korean entertainment companies are positioning themselves for a return to China after the two countries agreed to expand cultural exchanges, raising hopes that Beijing may loosen restrictions that have kept major K-pop acts out of the mainland for almost a decade.

中文摘要 韩国娱乐公司正为重返中国市场做准备。此前两国同意扩大文化交流,市场期待北京放松近十年来对韩国流行音乐团体赴华演出的限制。

A-Cap Insurer Asks to Defer Payouts to Clients Amid Deal Talks

Advantage Capital Holdings’ main life insurer, which has been disputing a state regulator’s concerns about its financial strength, is seeking to defer certain payments to customers while pursuing sale talks with a potential acquirer.

中文摘要 Advantage Capital Holdings旗下主要寿险公司因与州监管机构就财务实力产生争议,正寻求延迟向客户付款,同时与潜在收购方进行出售谈判。

Australian Dollar Erases 2026 Gain as Rate-Cut Bets Increase

Australia’s dollar wiped out its 2026 gain against the greenback after weak employment data reinforced bets that the central bank has ended its interest-rate hike cycle and will pivot to cuts next year.

中文摘要 澳元兑美元抹平2026年涨幅。疲弱就业数据强化了市场对澳洲联储结束加息周期、明年转向降息的预期。

Anthropic IPO Hands Korean Tycoon’s Firm $3 Billion Windfall

Anthropic PBC’s blockbuster initial public offering has delivered a multibillion-dollar windfall for billionaire Chey Tae-won’s empire, rewarding an early bet on the artificial intelligence startup by SK Telecom Co., South Korea’s largest mobile carrier.

中文摘要 Anthropic首次公开募股为韩国亿万富翁崔泰源的商业帝国带来数十亿美元回报。SK电信早期对这家AI初创公司的投资获得约30亿美元收益。

LNG Exports Through Hormuz Appear to Slow After More Attacks

The amount of liquefied natural gas tankers passing outbound through the Strait of Hormuz appears to have dropped over the last week, according to satellite and ship-tracking data, suggesting shipowners are being more cautious after a flurry of attacks.

中文摘要 卫星和船舶追踪数据显示,过去一周经霍尔木兹海峡出口的液化天然气船数量下降,船东在近期袭击事件后采取更谨慎态度。

该源今日无内容。