Python · ★ 9,645 · 🍴 636 · 📈 3,530 stars today
Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.
中文介绍 在内容送入 LLM 之前对工具输出、日志、文件和 RAG 片段进行压缩，可减少 60-95% 的 token 消耗，同时保持答案质量不变。提供库、代理和 MCP Server 三种接入方式，适合需要控制 AI 调用成本的开发者和企业，尤其适用于处理大量日志或长文档的 RAG 场景。
JavaScript · ★ 205,704 · 🍴 31,583 · 📈 2,141 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 面向 AI 编程智能体的性能优化框架，为 Claude Code、Codex、Opencode、Cursor 等工具提供技能、本能反应、记忆管理和安全增强能力。采用研究优先的开发理念，帮助开发者提升 AI 编程助手的实际执行效率与安全性，适合深度使用 AI 辅助编程工具的工程团队。
Go · ★ 35,393 · 🍴 415 · 📈 24 stars today
Find vulnerabilities, misconfigurations, secrets, SBOM in containers, Kubernetes, code repositories, clouds and more
中文介绍 Aqua Security 出品的全面安全扫描工具，支持检测容器镜像、Kubernetes 集群、代码仓库和云环境中的漏洞、配置错误、敏感信息及 SBOM。一条命令即可完成多维度安全审查，广泛用于 CI/CD 流水线和 DevSecOps 流程，是容器安全领域最受欢迎的开源工具之一。
Python · ★ 179,079 · 🍴 30,673 · 📈 1,735 stars today
The agent that grows with you
中文介绍 NousResearch 推出的可成长型 AI 智能体框架，随用户需求持续演进。基于 Hermes 系列模型构建，定位为通用 AI Agent 基础设施，适合需要长期运行、持续学习和个性化适配的智能体应用开发场景。
Python · ★ 142,828 · 🍴 9,762 · 📈 1,984 stars today
Python tool for converting files and office documents to Markdown.
中文介绍 微软开源的 Python 工具，可将 Word、Excel、PowerPoint 等 Office 文档及多种文件格式批量转换为 Markdown 文本。便于将传统文档接入 LLM 处理流程或知识库构建，适合需要文档预处理的 RAG 系统开发者和内容迁移场景。
Python · ★ 13,093 · 🍴 1,593 · 📈 719 stars today
Hermes WebUI: The best way to use Hermes Agent from the web or from your phone!
中文介绍 Hermes Agent 的 Web 前端界面，支持通过浏览器或手机端直接与 Hermes 智能体交互，提供友好的对话式操作体验。方便非技术用户无需配置命令行环境即可使用 Hermes Agent 的全部能力，适合团队共享部署场景。
Python · ★ 60,213 · 🍴 5,802 · 📈 1,067 stars today
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!
中文介绍 自适应 Web 爬虫框架，能够自动应对页面结构变化，从单次请求到大规模全站爬取均可胜任。内置反反爬策略，支持动态页面处理，适合数据采集工程师和需要长期稳定运行爬虫的业务场景，大幅降低因页面改版导致的维护成本。
Java · ★ 23,250 · 🍴 2,182 · 📈 570 stars today
PDF Parser for AI-ready data. Automate PDF accessibility. Open-source.
中文介绍 面向 AI 训练数据准备的开源 PDF 解析器，自动将 PDF 文档转换为结构化、可访问的数据格式，同时提升文档无障碍性。适合需要批量处理 PDF 文件用于大模型训练、知识库构建或 RAG 数据准备的开发者和数据工程师。
Python · ★ 51,922 · 🍴 32,642 · 📈 29 stars today
Odoo. Open Source Apps To Grow Your Business.
中文介绍 业界知名的开源企业管理套件，涵盖 CRM、ERP、电商、财务、项目管理、HR 等数十个模块，支持模块化按需部署。采用 Python 开发，拥有庞大的社区和丰富的第三方插件生态，适合中小企业及开发者构建定制化业务系统。
Python · ★ 8,937 · 🍴 1,106 · 📈 693 stars today
Talk to any LLM with hands-free voice interaction, voice interruption, and Live2D taking face running locally across platforms
中文介绍 本地跨平台运行的 LLM 虚拟主播框架，支持免提语音交互、语音打断和 Live2D 实时口型驱动，可对接任意大语言模型。适合 VTuber 创作者、AI 伴侣应用开发者及直播互动场景，所有计算本地完成，保护隐私且无需付费 API。
★ 348,995 · 🍴 83,153 · 📈 330 stars today
A complete computer science study plan to become a software engineer.
中文介绍 系统完整的计算机科学自学路线图，专为零基础或非科班背景人士备战大厂软件工程师面试而设计。涵盖数据结构、算法、系统设计、操作系统等核心知识点，附带学习资源和刷题建议，是 GitHub 上星标最多的面试备考项目之一。
Jupyter Notebook · ★ 18,879 · 🍴 2,070 · 📈 208 stars today
AirLLM 70B inference with single 4GB GPU
中文介绍 突破显存限制的大模型推理优化库，支持在单张 4GB 显卡上运行 70B 参数大模型推理。通过分层加载和内存管理技术实现低显存部署，使消费级 GPU 用户也能在本地运行超大规模语言模型，大幅降低硬件门槛。
TypeScript · ★ 25,153 · 🍴 2,209 · 📈 600 stars today
Memory engine and app that is extremely fast, scalable. The Memory API for the AI era.
中文介绍 专为 AI 时代设计的高性能记忆引擎，提供极速、可扩展的 Memory API，可为 AI 应用添加持久化记忆能力。支持存储、检索和管理用户上下文信息，适合需要跨会话记忆功能的 AI 助手、聊天机器人及个性化推荐系统开发者。
Python · ★ 9,892 · 🍴 2,000 · 📈 197 stars today
"Vibe-Trading: Your Personal Trading Agent"
中文介绍 香港大学数据科学实验室推出的个人化 AI 量化交易智能体，结合大语言模型与金融市场分析能力，辅助用户制定交易决策。适合量化研究者和金融科技开发者探索 LLM 在股票交易策略生成与执行中的应用潜力。
👍 6
Reasoning models improve accuracy through extended chains of thought, but their long outputs create a memory and compute bottleneck. KV cache eviction methods reduce this cost by evicting unimportant key-value pairs from the cache, yet they often yield worse accuracy than selection-based sparse atte
👍 21
World models and multimodal large language models (MLLMs) provide complementary capabilities for predicting future outcomes from static visual observations. World models can generate concrete visual rollouts of possible futures, while MLLMs can reason abstractly over questions, goals, and rules. How
👍 10
Test-time scaling improves the reasoning performance of large language models but incurs substantial cost in both total computation and latency. Existing adaptive sampling methods partially mitigate this issue by dynamically deciding when to stop sampling, yet they typically rely on heuristic rules
👍 4
A core goal of computational social science is to discover interpretable differences in how language varies across outcomes of interest, such as political affiliation or instructional quality. Recent LLM-based hypothesis generation methods describe such differences in natural language, but select fo
👍 3
Real-time vision demands models that are accurate, efficient, and simple to deploy across diverse hardware. The YOLO family has become widely deployed for this reason, yet most YOLO detectors still rely on non-maximum suppression at inference, carry heavy detection heads due to Distribution Focal Lo
👍 26
Test-time scaling is a powerful approach to obtain better reasoning in large language models, but it becomes memory-bottlenecked during long-horizon decoding, as the KV-cache grows. KV-cache quantization can help improve this, but current methods are evaluated under prefill-like settings and errors
👍 12
Modern generative models possess a deep understanding of visual content, yet training them for image editing typically requires massive datasets of paired examples. This limits scalability, especially for video editing where collecting paired data is prohibitively expensive. We propose Bootstrap You
👍 9
We introduce PaddleOCR-VL-1.6, an upgraded compact document parsing model built upon PaddleOCR-VL-1.5. Although PaddleOCR-VL-1.5 establishes a strong 0.9B baseline, its remaining errors concentrate in under-optimized regions where model behavior is unstable, data coverage is sparse, or supervision i
👍 13
As autonomous vehicle capabilities advance, the safe evaluation of driving policies in long-tail scenarios remains a critical bottleneck. In closed-loop simulation, the driving policy model actively interacts with the environment, where its actions dynamically update the simulator state and directly
👍 14
The past few decades have witnessed significant advances in the design of machine learning algorithms, from early studies on task-specific shallow models to more general deep Large Language Models (LLMs). Despite showing promising results in tasks that require instant prediction or in-context learni
👍 2
The KV-cache is the right memory for datacenters but the wrong memory for robots. Datacenter inference batches many short requests and resets them, amortizing an attention cache across a crowd. Embodied agents instead run one long, non-resetting episode on bandwidth-limited edge hardware, where high
👍 11
Personalization is a crucial capability of modern language agents. However, current research primarily positions personalized agents as passive responders to user preferences, limiting their ability to interact with users and provide suggestions or guidance proactively. To systematically evaluate su
👍 10
Auto-harness systems such as A-Evolve, GEPA, and Meta-Harness improve LLM agents by optimizing prompts, skills, tools, memories, and supporting infrastructure from execution feedback, but they are typically evaluated on fixed offline benchmarks. Real deployments instead present open-ended task strea
👍 16
Reinforcement learning (RL) for visual reasoning needs scalable, verifiable, and controllable training signals. Existing visual RL post-training trains on static curated datasets, with fixed image-question-answer samples bounded by their collection budget. In this work, we introduce TRON (Targeted,
👍 0
WALL-WM is a World Action Model that shifts video-action learning from chunk-centric optimization to event-grounded Vision-Language-Action pretraining, using semantically coherent action events as the atomic unit of learning. Existing WAMs commonly initialize from multimodal or video foundation mode
👍 1
Recent multimodal large language models have demonstrated strong reasoning ability, yet their reliability as automated evaluators remains limited by a critical weakness: when visual evidence conflicts with textual cues, MLLM judges tend to reward plausible narratives over perceptually correct answer
👍 20
Autonomous agents are increasingly expected to support end-to-end medical-AI research workflows, moving beyond isolated prediction tasks or short-form clinical question answering. However, existing medical agent benchmarks primarily evaluate final outputs, providing limited visibility into agent beh
👍 24
Reinforcement learning (RL) post-training improves large language models (LLMs) on individual domains such as mathematical reasoning, code generation, question answering, and creative writing (CW), but training on one domain often degrades performance on others. Existing explanations based on catast
👍 8
Instruction tuning aligns large language models, including multimodal ones, with diverse user intents, but scaling to heterogeneous mixtures is hindered by gradient interference and bandwidth-heavy synchronization. We ask whether these two bottlenecks can be addressed jointly by training parts of th
👍 8
Embodied visual navigation, where an agent perceives a complex environment and acts to reach a goal from raw sensory input, underpins a wide range of applications such as household service robotics, assistive robotics, and large-scale autonomous exploration. However, recent attempts to unify vision-
👍 3
Agent skills extend AI agents with reusable instructions, tools, scripts, references, and workflows, establishing a security boundary distinct from both model safety and traditional package-malware detection. ClawHub Security Signals is a sanitized dataset of 67,453 latest public OpenClaw skill vers
👍 7
On-Policy Distillation (OPD) trains a student model on its own generative trajectories under dense token-level feedback from a stronger teacher, mitigating both the off-policy distribution shift of Supervised Fine-Tuning (SFT) and the sparse credit assignment of Reinforcement Learning (RL). However,
👍 72
Recent progress in the development of language models has been defined by scale, with each generation absorbing more of the world's knowledge into its weights. However, many practical applications benefit more from robust reasoning than from extensive parametric knowledge. In this setting, task-spec
👍 4
Large Language Models have demonstrated remarkable progress in general-purpose capabilities and can achieve strong performance in specific domains through fine-tuning on domain-specific data. However, acquiring high-quality data for target domains remains a significant challenge. Existing data synth
👍 19
Mid-training has become an important stage in modern LLM development, using large-scale curated mixtures to strengthen capabilities before final post-training. Its data selection problem is distinct: the data are optimized under a pretraining-style objective at near-pretraining scale, but are curate
👍 4
Finite element analysis (FEA) is the most important numerical approach for solid mechanics. Challenges of FEA include a steep learning curve for entry-level users and potential false simulations due to incorrect definitions of key simulation components, such as boundary conditions, load cases, and s
👍 8
Long chain-of-thought (CoT) traces are widely used as supervision for reasoning-oriented LLM SFT, yet answer-correct traces can still lead to markedly different fine-tuning outcomes. We study post-conclusion continuation in answer-correct long-CoT data: a continuation where the answer appears suffic
👍 1
Linear probes trained on LLM activations are increasingly proposed as deception-detection metrics, yet report AUROC exceeding 0.96 on clean benchmarks while collapsing under distributional shift. This paper systematically pressure-tests probe-based metrics across the Gemma 3 model family (1B-27B par
👍 7
Current music similarity models typically compute a single, monolithic score, entangling distinct musical dimensions like melody, rhythm, and timbre. This limits user control and interpretability, making it impossible to execute nuanced queries. We introduce MERIT, a framework for learning disentang
👍 40
Identifying which brain regions represent a visual concept in the human brain is a central challenge in neuroscience. Existing approaches have localized coarse functional regions (e.g., faces, places) through activation maximization, identifying regions that activate strongly for a target concept re
@elpresidank · 116 粉丝 · 2.9M 阅 · 543 赞 · 35 转
Most AI agent memory is built on embeddings. And there's now a proof that this entire class of system is going to forget what you stored in it — and confidently make up things you never stored at all.
中文介绍 基于嵌入向量的 AI 智能体记忆系统存在根本性缺陷：有数学证明表明此类系统会遗忘已存储内容，并自信地「捏造」从未存储过的信息。作者从拓扑学视角分析记忆失效根因，提出用「上下文结构」而非纯向量相似度来组织智能体记忆，是对当前主流 agent 记忆架构的深度批判与反思。
@1salman · 363 粉丝 · 2.0M 阅 · 682 赞 · 45 转
Everyone keeps asking whether AI favors specialists or generalists. I think that is the wrong question. AI does not pick a side. It changes the tradeoff. The old world forced a choice. You could go
中文介绍 作者认为「AI 究竟利好专才还是通才」是个伪命题。AI 并不选边，而是根本改变了专精与广博之间的权衡关系：过去必须二选一，现在 AI 让人可以按需切换深度与广度，传统职业路径的底层逻辑正在被重写。
@prukalpa · 23.1K 粉丝 · 583.2K 阅 · 506 赞 · 80 转
A field guide to what it is, what it is not, and where it fits in your AI architecture. I have had some version of the same conversation with a CIO almost every day this year. Their team has read
中文介绍 针对企业 AI 架构中「企业上下文层（Enterprise Context Layer）」的概念混乱问题，作者以每天与 CIO 对话的一线经验，整理出这一层到底是什么、不是什么、以及在 AI 技术栈中的定位。适合正在规划企业 AI 基础设施的技术决策者参考。
@theonejvo · 22.1K 粉丝 · 504.3K 阅 · 861 赞 · 1 转
Over the past year, @pewdiepie, has been turning into one of the most visible champions of private, self-hosted computing, and it has been a genuine pleasure to watch. What began in late 2025 as an
中文介绍 作者对 PewDiePie 自托管私有计算项目中的 AI agent 框架（harness）发起提示注入攻击实验：通过一个伪装成儿童视频网站的恶意页面成功入侵其 agent，随后协助其修补漏洞。全程展示了 agent harness 的真实安全风险与防御思路。
@garrytan · 853.3K 粉丝 · 180.6K 阅 · 503 赞 · 43 转
In January I got back into coding and I built Garry's List. Over five hundred thousand lines of Rails and the tests to police it. I was proud of it. I shouldn't have been. The thing worth being proud
中文介绍 Garry Tan 反思自己用 AI 写了 50 万行 Rails 代码的经历，认为「堆代码量」是错误的骄傲来源。真正值得骄傲的是产品价值，而非代码规模。他警告开发者不要把 AI agent 当流水线工厂来用，避免陷入「用 AI 造更大的技术债」的陷阱。
@mem0ai · 17.6K 粉丝 · 82.8K 阅 · 520 赞 · 60 转
Agent harnesses are where AI software actually runs. Cursor, Devin, Claude Code, Codex: these environments handle context, orchestrate tools, coordinate agents, and increasingly, manage memory. The
中文介绍 mem0 团队梳理了主流 agent harness（Cursor、Devin、Claude Code、Codex）在记忆管理上的现状与差异。这些环境负责上下文管理、工具编排与多 agent 协调，记忆能力正成为关键竞争维度。文章提供了各平台记忆机制的横向对比视角。
@trq212 · 263.1K 粉丝 · 75.7K 阅 · 542 赞 · 36 转
Last week, we released dynamic workflows in Claude Code. Claude can now write its own harness on the fly, custom-built for the task at hand. While the default Claude Code harness is built for coding,
中文介绍 Claude Code 新推出「动态工作流」功能：Claude 可以在执行任务时即时编写专属 harness，不再局限于默认的编程场景框架。作者介绍了该功能的设计理念与使用场景，意味着 Claude Code 向通用 agent 运行时又迈进一步。
@drfeifei · 738.0K 粉丝 · 72.2K 阅 · 699 赞 · 144 转
“The world is everything that is the case.” — Ludwig Wittgenstein, Tractatus Logico-Philosophicus, 1921 The world is not made of words. In an earlier essay, we argued that spatial intelligence is AI’s
中文介绍 李飞飞发表长文，构建「世界模型的功能性分类体系」。她以维特根斯坦名句开篇，论证世界不由语言构成，延续其空间智能议题：AI 理解世界需要超越文字的多层次世界模型。文章属于其 AI 基础认知框架系列的深度理论探讨。
@mvanhorn · 27.6K 粉丝 · 54.5K 阅 · 545 赞 · 44 转
Three months ago I posted "Every Claude Code Hack I Know." It hit 913K views. @kevinrose had asked what IDE to use, and my answer was: "No IDE. Just plan.md files and voice." This used to be called
中文介绍 作者 3 个月前发布「Claude Code 实战技巧」获 91.3 万浏览，此次更新为 2026 年 6 月版「Agentic 工程全套技巧」。核心理念依旧：不用 IDE，只用 plan.md 文件加语音输入驱动 agent，并补充了大量新积累的工程实践经验。
@0xJeff · 80.5K 粉丝 · 47.0K 阅 · 512 赞 · 45 转
It’s been a little more than 2 months now and Hermes has taught me so many different things. One of the key things is that Agents like Hermes or OpenClaw don’t fail on intelligence, they tend to fail
中文介绍 作者使用 AI 分析师 agent「Hermes」60 天、跑了 6 套工作流后的复盘总结。核心发现：Hermes、OpenClaw 等 agent 失败的根源不在智能不足，而在于其他环节——提示词结构、工具调用边界、任务拆解质量等执行层问题才是瓶颈。
@OptimaiNetwork · 98.8K 粉丝 · 5.7K 阅 · 542 赞 · 110 转
May was an important month for OptimAI because it marked a shift from proving the product to expanding the platform. The broader market is moving in the same direction: AI is no longer just about
中文介绍 OptimAI Network 发布 5 月月报与 6 月规划。5 月完成了从「验证产品」到「扩展平台」的阶段性转变，方向是构建「agent 原生互联网」基础设施。内容偏项目进展通报，反映当前市场向 AI agent 基础层快速迁移的趋势。
Verified Generation and Compounding Intelligence
中文介绍 Latent Space播客邀请Axiom Math联合创始人Carina Hong，探讨如何将人工智能从非正式应用推向规范化规模化，核心议题包括"验证生成"（Verified Generation）与"复利智能"（Compounding Intelligence）两大方向，展示数学推理领域AI的前沿发展路径。
The legendary Microsoft CEO makes his first Latent Space appearance!
中文介绍 微软CEO萨提亚·纳德拉首次亮相Latent Space播客，以No Priors与Latent Space联合特别节目形式，在微软Build开发者大会期间进行对话，分享其对AI发展及微软战略的最新观点。
GPT-Rosalind advances life sciences research with enhanced biological reasoning, medicinal chemistry expertise, genomics analysis, and experimental workflow capabilities.
中文介绍 OpenAI推出GPT-Rosalind新版本，面向生命科学研究领域，新增增强型生物推理、药物化学专业知识、基因组学分析及实验工作流管理等核心能力，旨在加速生命科学研究进程。
中文介绍 Hugging Face博客发布文章，探讨直接偏好优化（DPO）技术在聊天机器人以外更广泛场景的应用潜力与实践方向，具体技术细节及应用案例有待进一步披露。
See how Wasmer used Codex with GPT-5.5 to build a Node.js runtime for the edge, accelerating development 10x to 20x and shipping in weeks instead of months.
中文介绍 边缘计算公司Wasmer借助OpenAI Codex（基于GPT-5.5）构建面向边缘场景的Node.js运行时，开发效率提升10至20倍，原本需要数月完成的项目压缩至数周内交付，成为AI辅助开发的典型案例。
OpenAI outlines its public policy agenda for AI, including safety, youth protection, workforce transition, and global standards to ensure AI benefits society.
中文介绍 OpenAI正式发布公共政策议程，涵盖AI安全、青少年保护、劳动力转型及全球标准制定四大核心议题，明确公司在推动AI技术惠及社会方面的政策立场与优先事项。
OpenAI outlines a blueprint for U.S. governance of frontier AI, proposing a federal framework for safety, resilience, and national security.
中文介绍 OpenAI发布前沿AI民主治理蓝图，面向美国政府提出联邦层面的监管框架建议，重点涵盖AI安全、系统韧性与国家安全三大维度，呼吁建立统一的前沿AI治理机制。
Microsoft Build recap, and new MAI model technical details
中文介绍 微软Build大会发布MAI-Thinking-1及MAI系列模型，Latent Space对本届大会进行全面回顾，并披露MAI模型的技术架构细节，标志着微软在自研大模型领域的重要进展。
中文介绍 Hugging Face博客介绍如何为Reachy Mini机器人集成MCP工具，探索将模型控制协议（MCP）应用于实体机器人平台的实践路径，具体实现细节及性能数据有待查阅原文。
中文介绍 TLDR AI日报（2026年6月3日）聚焦三大话题：OpenAI推出Codex Sites网站生成功能、微软Build大会发布多款新模型，以及市场对Anthropic服务定价上涨的争议与用户反弹。
GitHub pioneered the modern AI coding era with Copilot, and the resulting explosion in agentic coding has led to notable strains on the most popular developer platform in the world. Here's the plan.
中文介绍 GitHub首席产品官Kyle Daigle在Latent Space播客分享GitHub的AI Agent战略规划。GitHub凭借Copilot开创现代AI编程时代，但Agent化编程浪潮也给这一全球最大开发者平台带来显著的基础设施压力，文章详述应对计划。
中文介绍 H Company在Hugging Face发布Holo3.1模型，定位为快速、本地化运行的计算机操作Agent，支持在本地设备上执行计算机使用任务，强调低延迟与隐私保护特性。
Travelers built an AI-powered Claim Assistant with OpenAI to guide customers through filing claims, provide 24/7 support, and scale operations during peak demand.
中文介绍 美国保险巨头Travelers与OpenAI合作，在全国范围内部署AI驱动的理赔助手（Claim Assistant），为客户提供全天候24小时理赔引导服务，并在理赔高峰期实现业务弹性扩容，覆盖全美各地运营。
The global health care sector is under increasing strain. Decades of chronic underinvestment and constraints in recruitment have coincided with a surge in demand for services for aging populations. Gaps in provision are already taking a toll, with fragmented access to care and high rates of stress a
中文介绍 麻省理工科技评论探讨Agentic AI在全球医疗体系改造中的潜力。受长期投入不足、人员招募困难及老龄化人口需求激增三重压力影响，全球医疗体系持续承压，文章分析如何借助自主AI Agent弥合医疗资源碎片化缺口，重塑医患关系。
Discover new Codex plugins, sites, and annotations that help analysts, marketers, designers, investors, and other teams get more done with AI.
中文介绍 OpenAI扩展Codex应用生态，推出面向多职能角色的插件、Codex Sites及代码注释等新功能，覆盖分析师、营销人员、设计师、投资者等非技术岗位，旨在将AI编程能力延伸至各类业务工作流。
本期技术面呈现明显的「结构分化」格局。美股方面，SPY 报 754.24，RSI14=67.9，多头排列完整（SMA20=744.83 / SMA50=709.82 / SMA200=682.87），距 52 周高仅 -0.81%，整体动能维持偏强；QQQ 更进一步触发 RSI 超买（77.3），MACD 金叉叠加多头排列，短期动量较全市场最为强劲。但个股层面出现分化：MSFT 单日跌 -3.17%，NVDA 跌 -3.62%，GOOGL 5 日累跌 -7.67%，RSI 降至 38.8，指数强势掩盖了结构性弱势。加密资产整体遭受重压：BTC 报 64,221，5 日跌幅 -12.47%，ETH 5 日跌 -9.81%，SOL 5 日跌 -12.2%，三者 RSI 均跌入超卖区（18.7 / 20.0 / 23.1），空头排列完整；加密恐慌贪婪指数低至 12（极度恐慌），加密总市值 24h 缩减 2.94% 至 2.32T USD，BTC 主导率 55.3% 显示资金向主流币集中防御。中概股全线承压，BABA / PDD / JD / 腾讯均呈空头排列，RSI 普遍在 36-52 区间徘徊，腾讯虽 5 日涨 +7.37% 并触发 MACD 金叉，但仍处空头排列下的超跌反弹形态。商品与外汇方面，美元指数 DXY 报 99.46，多头排列，RSI=59.5，动能偏强；WTI 原油 5 日涨 +7.52%，但 MACD 仍为负值；黄金期货 RSI=41.4，价格低于 SMA20/SMA50，短期偏弱。10 年期美债收益率升至 4.49%，RSI=53.5，多头排列，对权益市场形成一定压力。综合来看，美股指数技术面偏强但个股分化加剧，加密资产技术面显著偏弱，商品外汇分化，中概承压。
QQQ 当前报 744.21，5 日涨幅 +2.02%，距 52 周高仅 -0.59%。RSI14 达 77.3，处于超买区间；MACD 为 21.7618，略高于信号线 21.6286，呈金叉状态。三条均线呈多头排列（SMA20=719.26 > SMA50=662.33 > SMA200=620.32），价格站在所有均线之上。技术动量强劲，但 RSI 超买读数提示短期可能面临动能衰竭风险。
BTC 当前报 64,221.96，1 日跌幅 -3.72%，5 日跌幅 -12.47%，距 52 周高达 -49.11%。RSI14 仅 18.7，深度超卖；MACD 为 -2470.04，大幅低于信号线 -1266.79，空头动能扩张。价格位于 SMA20（74,820）、SMA50（76,887）、SMA200（79,085）全部下方，呈空头排列。辅以加密恐慌贪婪指数仅 12（极度恐慌）、加密总市值 24h 下跌 2.94%，技术面与市场情绪均指向持续下行压力。
ETH 报 1,814.67，5 日跌幅 -9.81%，距 52 周高 -63.37%。RSI14 为 20，逼近极度超卖边界；MACD -88.88 持续低于信号线 -68.23，绿柱扩张显示空头加速。三均线 SMA20=2,061、SMA50=2,217、SMA200=2,478 全线压制价格，空头排列完整。ETH 在加密总市值中占比 9.4%，与整体加密市场同步走弱，技术结构未见企稳迹象。
GOOGL 报 358.99，5 日跌幅 -7.67%，距 52 周高 -12.14%。RSI14 降至 38.8，接近超卖边界；MACD 3.34 但信号线已升至 9.53，MACD 柱快速收缩，动量明显减弱。价格已大幅跌破 SMA20（388.14），尽管整体仍为多头排列，短期走势已背离均线方向，短线技术状态趋于疲软。
META 报 622.98，当日大涨 +4.24%，但 5 日仍跌 -1.93%，距 52 周高 -21.76%。RSI14 为 51.5，位于中轴附近；MACD -3.08 刚上穿信号线 -3.80 形成金叉，直方图由负转正但绝对值尚小，动能方向切换初现但力度有限。SMA50=618.87 与当前价接近，SMA200=663.94 仍压制价格上方，趋势方向尚不明朗，技术面偏中性观望。
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
本报告仅基于公开行情数据计算的技术指标进行客观描述，过去走势不代表未来表现，所有指标读数及形态描述均不构成任何投资建议或买卖推荐，仅供技术指标解读参考。投资者应结合自身风险承受能力独立决策。
The U.S.-brokered agreement called for “a complete cessation” of fire from Hezbollah, an Iranian-backed militia that operates in southern Lebanon. A previous cease-fire has largely been ignored.
中文摘要 以色列与黎巴嫩在美国斡旋下达成协议，同意续签停火协议，协议要求真主党（伊朗支持的黎巴嫩武装组织）"完全停火"。此前一份停火协议已基本遭到双方忽视。
In the 215-208 vote, four Republicans joined Democrats to pass the measure, which had failed in three previous attempts.
中文摘要 美国众议院以215票对208票通过决议，要求叫停对伊朗的战争行动，此举被视为对特朗普政策的公开抵制。投票中有4名共和党议员与民主党联合支持该决议，该决议此前已三度在众议院表决失败。
Statement with US after Israel-Lebanon talks in Washington says agreement will create ‘pilot’ security zones in Lebanon from which Hezbollah banned One killed and 63 hurt in Iran attack on Kuwait airport as Trump says ceasefire talks ongoing The Kuwaiti defence ministry said it intercepted 13 ballis
中文摘要 以色列与黎巴嫩在华盛顿会谈后达成续签停火协议，协议将在黎巴嫩建立禁止真主党进入的"试点"安全区。与此同时，伊朗对科威特机场发动袭击，造成1人死亡、63人受伤，特朗普表示停火谈判仍在继续。美国众议院亦通过决议，就伊朗战争问题向特朗普施压。
Greens and Coalition oppose discretionary powers that would allow the treasurer to make rule changes to CGT and negative gearing after the legislation passes. Follow today’s news live Get our breaking news email, free app or daily news podcast Jim Chalmers’ budget bill will face debate and a vote in
中文摘要 澳大利亚财长吉姆·查尔默斯提出的预算法案在议会遭遇阻力，绿党与联合党联合反对其中赋予财长对资本利得税及负扣税规则进行自由裁量修改的条款，绿党形容此为"亨利八世式权力"，要求对相关新税法进行严格审查。
Activists have tried to erect a statue of Marwan Barghouti in London’s Parliament Square, but police intervened.
中文摘要 伦敦活动人士尝试在议会广场树立在押巴勒斯坦领导人马尔万·巴尔古提的雕像，但遭到警察介入阻止。
中文摘要 该页面为《纽约时报》关于伊朗战争及以色列、黎巴嫩局势的实时更新专题，具体内容暂无摘要信息。
It was the latest attempt by President Trump to flip the narrative on a conflict that he once said would lead to quick and decisive victory.
中文摘要 特朗普公开表示，对伊战争对美国而言"不是什么大事"，此番言论是他持续尝试转变外界叙事的最新举动。此前他曾宣称此次冲突将迅速取得决定性胜利，但目前局势仍未如预期发展。
Canada is the first G7 country to approve the lower cost weight-loss drugs, which are rolling out in pharmacies this month.
中文摘要 加拿大成为七国集团（G7）中首个批准仿制版司美格鲁肽（Ozempic）减重药物的国家，相关低价仿制药本月已开始在药房推出。美国目前尚未批准仿制版本。
Despite long-established procedures for bringing Americans home for monitoring and treatment, the Trump administration has not said that it will allow those at risk of Ebola back into the country.
中文摘要 埃博拉疫情持续扩散，但特朗普政府迄今未公布应对方案。尽管美国历来有成熟程序用于接回存在埃博拉感染风险的本国公民进行监测和治疗，当局目前仍未表态是否允许相关风险人员返回美国。
High fuel prices are strengthening the case for EV adoption in the world's third largest auto market.
中文摘要 高油价正推动更多印度消费者转向电动汽车。印度是全球第三大汽车市场，燃油成本高企正进一步强化电动汽车的经济性优势，但充电基础设施不足等挑战依然存在。
Trump says he wants to keep talks on the conflict in Lebanon separate from negotiations over the war with Iran.
中文摘要 特朗普表示希望将黎巴嫩问题谈判与对伊朗战争谈判分开处理，主张两条线索独立推进，避免相互捆绑。
Kuwait has released surveillance footage of what it says shows an Iranian drone strike on Kuwait International Airport.
中文摘要 科威特公布科威特国际机场遭袭的闭路电视监控录像，画面显示伊朗无人机对机场实施打击的过程。此次袭击造成人员伤亡，科威特称视频证实了伊朗的攻击行为。
President Trump appears ready to lend a hand to a new right-wing ally ahead of a crucial election this year that will be the biggest test for Latin America’s left.
中文摘要 特朗普再度介入巴西政治，据报准备为新右翼盟友提供支持，以影响今年至关重要的巴西大选。此次选举被视为对拉丁美洲左翼势力的最大考验。
The administration has settled on a more legally and politically durable way to impose tariffs, but some say the focus on forced labor laws is merely a pretext for protectionism.
中文摘要 特朗普政府以强迫劳动相关法律为由，为现行关税政策寻求新的法律与政治依据。批评人士认为，援引强迫劳动法规不过是推行贸易保护主义的借口，而非实质性理由。
Real Madrid president Florentino Perez says he will bring Jose Mourinho back as manager if he wins Sunday's election.
Shares in chip company fall as much as 15% in after-hours trading
中文摘要 芯片巨头博通（Broadcom）因营收预测令市场失望，盘后股价一度暴跌15%，市值蒸发逾3000亿美元。此次业绩指引未达投资者预期，引发市场对其增长前景的担忧。
The Indonesian rupiah is closing in on a key psychological level, putting market watchers on watch for a stronger response from the central bank.
中文摘要 印度尼西亚卢比汇率逼近18000关口这一关键心理水平，市场人士密切关注印尼央行是否会采取更强力的干预措施以稳定汇率，防止进一步贬值。
Elon Musk’s SpaceX is poised to have the largest stock-market debut in history when it goes public later in June. SpaceX is aiming to raise $75 billion in the initial public offering, more than twice the previous record holder.
中文摘要 埃隆·马斯克旗下SpaceX计划于6月晚些时候上市，目标募资750亿美元，有望成为史上规模最大的IPO，超出此前纪录持有者逾一倍。
Gil Luria, DA Davidson Head of Technology Research, discusses the outlook for Broadcom as the Nvidia alternative releases its latest quarterly data. (Source: Bloomberg)
中文摘要 DA Davidson科技研究主管Gil Luria就博通最新季度业绩数据发表分析，探讨博通作为英伟达替代选项的市场前景及其与同业竞争者的比较表现。
Gold held a decline after renewed hostilities between the US and Iran jeopardized talks to end the war, keeping energy prices and inflation risks elevated.
中文摘要 黄金延续跌势，美伊之间爆发新一轮冲突，令两国和平谈判前景蒙上阴影，能源价格承压上行，通胀风险上升，投资者持观望态度。
The more than $90bn hedge fund tells investors it will take longer to retrieve their money from two flagship funds
中文摘要 管理规模逾900亿美元的对冲基金DE Shaw通知投资者，旗下两只主力基金将延长资金锁定期至最长四年，投资者赎回资金所需时间将比预期更长。
Elon Musk’s space exploration company set a target price for buyers earlier than expected.
中文摘要 SpaceX提前披露IPO目标定价，公司估值设定为1.75万亿美元，较原定时间表提前向投资者公布，为即将到来的股市上市做准备。
UK-based investors pulled money from cash funds and piled into bonds last month, enticed by surging yields, according to fund flow data compiled by Calastone.
中文摘要 据Calastone基金流向数据显示，受债券收益率大幅攀升吸引，上月英国投资者纷纷从现金基金撤资，转而大举买入债券，资金配置出现明显转移。
The rising popularity of lab-grown diamonds heaps pressure on those hunting for the natural gems.
中文摘要 实验室培育钻石日益普及，对天然钻石开采行业形成压力。天然钻石矿工面临需求下滑与价格竞争的双重挑战，行业生存困境加剧。
The Commodity Futures Trading Commission is ending an almost 30-year-old policy that forbade companies and individuals from airing their grievances or denying wrongdoing after they settled enforcement cases with the derivatives regulator.
中文摘要 美国商品期货交易委员会（CFTC）宣布废除实施近30年的「禁言规则」，此前该规则禁止与CFTC达成执法和解的公司或个人公开否认不当行为或表达异议。
A $2.25 billion leveraged loan for McKesson Corp.’s medical-surgical unit priced on Wednesday, ahead of its planned initial public offering.
中文摘要 麦克森公司（McKesson）旗下医疗外科业务部门已于周三完成22.5亿美元杠杆贷款定价，为其计划中的首次公开募股铺路。
Rocket and AI group is seeking to raise up to $86bn in biggest Wall Street debut of all time
中文摘要 SpaceX向投资者推介IPO，估值高达1.8万亿美元，计划募资最多860亿美元，有望创下华尔街史上规模最大IPO纪录。该公司业务涵盖火箭发射与人工智能领域。
19 回复 · 程序员 节点
7 回复 · 程序员 节点
7 回复 · 程序员 节点
7 回复 · 程序员 节点
15 回复 · 程序员 节点
11 回复 · Apple 节点
14 回复 · 程序员 节点
8 回复 · Apple 节点
23 回复 · Apple 节点
17 回复 · Apple 节点
首字效果： codex公益站：https://new.sharedchat.cc/ 群内佬友评价： 一个人一个号不要重复注册，会风控 72 个帖子 - 61 位参与者 阅读完整话题
本帖使用社区公益推广，符合推广要求。我申明并遵循社区要求的以下内容： 我的项目是免费使用的，无收费（变相收费、赞助）部分： 是 我的帖子已经打上 公益推广 标签： 是 我的项目属于个人项目，与公司或商业机构无关： 是 我的项目不存在QQ、TG等群组引流： 是 我的项目不存在非运营必要的网站引流： 是 我的项目不存在为他人推广、AFF： 是 我的项目无关联的商业项目： 是 我的站点存在登录，并已接入 LINUX DO Connect： 是 我帖子内的项目介绍，AI生成、润色内容部分已截图发出： 是 以上选择我承诺是永久有效的，接受社区和佬友监督： 是 以下为项目介绍正文内容，AI生成、润色内容已
本帖使用社区开源推广，符合推广要求。我申明并遵循社区要求的以下内容： 我的帖子已经打上 开源推广 标签： 是 我的开源项目完整开源，无未开源部分： 是 我的开源项目已链接认可 LINUX DO 社区： 是 我帖子内的项目介绍，AI生成、润色内容部分已截图发出： 是 以上选择我承诺是永久有效的，接受社区和佬友监督： 是 以下为项目介绍正文内容，AI生成、润色内容已使用截图方式发出 github.com GitHub - fish2018/webhtv: WebHomeTV 基于FongMi二次开发，增强了 WebHome 自定义首页、App... WebHomeTV 基于FongMi二次开发，增
greasyfork.org Linux.do Boost 增强 选中 boost 后增加 @ 按钮，一键打开回复框并 @ 该 boost 用户 27 个帖子 - 14 位参与者 阅读完整话题
给它临时开的权限模式，让我帮我排查下什么导致电脑卡顿，查完以后说让我重启，重启以后直接不见了。。。 18 个帖子 - 18 位参与者 阅读完整话题
昨天看到佬友写的帖子 【π】关于pi的完善之路，package生态的折腾之路 感觉真心不错，刚睡醒，心血来潮也来分享一下自己近期折腾 pi 扩展的经验。 先给误入此帖 (bushi) 的佬介绍下， Pi 是一个轻量实现的 harness, 跟 claude code / codex / opencode 这些工具比起来绝对是简易甚至可以称得上是毛坯房了，很多功能作者都刻意避开，这也让它的 system prompt token 开销很小。 虽然 Pi 这类 agent 可应用的领域不止 coding，但我目前主要还是用于 coding。本贴推荐的都是一些基础插件，coding 场景还是都能用得
Map of Intl 是一个属于两个人的旅行记录应用。 最开始只是想解决一个很简单的问题： 旅行结束后，照片越来越多，但回忆却越来越难整理。 于是我尝试用 AI 和 Vibe Coding 的方式，做了一个能够记录共同生活的小程序。 它将地图、时间线、旅行计划与回忆整理融合在一起，把一起去过的城市、走过的路线、拍下的照片和重要纪念日串联成完整的故事。 随着记录越来越多，地图上的足迹会不断增长，时间线也会越来越长。 后来发现，它记录的不只是旅行。 而是两个人一起走过的生活。 希望多年以后再次打开时，依然能够看见那些共同经历过的风景和时光。 可自由编辑路线 可查看旅行日记 接入高德地图API 3
这是你不仁别怪我不义。我是你忠实粉丝，就这样回报我。你彻底成功的让我对你失望透顶 我这么纯都不行 既然你很喜欢发验证码是吧，那我就写个脚本让你全国发，不要停 36 个帖子 - 30 位参与者 阅读完整话题
如果有特别想关注的，可以在提示词中说的更清楚一些，codex拿到的数据还是蛮完善的，我现在就是每天看一下，太忙，最近都空仓着，没咋关注这些。 提示词（可根据需求自行修改，我样式上做了一些细微的调整，不好总结出来了[token 我没算过，我开的pro20x 会员，这个具体会跑多少token，我没细看过，记得注意哈，万一撑不住就不好了]）。 github链接：GitHub - ningzaichun/ai_stock_daily_report: codex ai stock daily report · GitHub 大家感兴趣可以看看，完整提示词和使用的案例都在里面。 49 个帖子 - 28 位
gpt 恢复供应 大家尽量不要给我发私信，有问题评论就行，无力回复，太多了 不要以任何理由索要额度/寻求解封 用不了就是号池在维护 不要去星辰的售后群去问我的客服为啥君の公益用不了，没有别的意思，就是客服可能会比较崩溃 我不会建任何有关君の公益的qq群/tg群，我也不喜欢看到有人以我的名义建群 君の公益唯一的解释权在我这里，我不喜欢看到有人代替我解释什么，希望大家尽量避免 签到额度调整为每天52刀 已经安排佬友帮我查分发了，已经封了一批，后面不定期也会安排检查 341 个帖子 - 294 位参与者 阅读完整话题