每日简报

2026-06-09

← 历史归档

mvanhorn/last30days-skill

Python · ★ 34,780 · 🍴 2,846 · 📈 3,558 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及全网搜索任意主题,并综合生成有依据的摘要。适用于需要快速聚合多平台最新信息进行洞察的分析师、研究人员或市场观察者。

RyanCodrai/turbovec

Python · ★ 9,006 · 🍴 817 · 📈 1,729 stars today

A vector index built on TurboQuant, written in Rust with Python bindings

中文介绍 基于TurboQuant构建的高性能向量索引,采用Rust开发并提供了Python绑定。专为需要快速向量检索的场景设计,例如RAG系统、语义搜索或推荐引擎,适合追求低延迟和高吞吐量的AI工程师。

google/skills

Python · ★ 12,488 · 🍴 970 · 📈 461 stars today

Agent Skills for Google products and technologies

中文介绍 Google官方发布的Agent技能库,涵盖旗下产品与技术的交互能力。开发者可利用这些技能快速构建能与Google搜索、日历、地图等服务协作的智能代理,简化集成流程。

refactoringhq/tolaria

TypeScript · ★ 13,639 · 🍴 954 · 📈 651 stars today

Desktop app to manage markdown knowledge bases

中文介绍 一款桌面应用,专注于管理Markdown格式的知识库。支持本地化存储、快速检索与笔记组织,适合习惯用Markdown写作的开发者、研究者或知识工作者进行个人知识管理。

Panniantong/Agent-Reach

Python · ★ 24,327 · 🍴 2,042 · 📈 679 stars today

Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.

中文介绍 通过CLI赋予AI代理全网读取与搜索能力,支持Twitter、Reddit、YouTube、GitHub、B站、小红书等平台,无需API费用。适合需要让AI自动爬取和分析多源网络信息的开发者和自动化场景。

danielmiessler/Personal_AI_Infrastructure

TypeScript · ★ 15,471 · 🍴 2,162 · 📈 62 stars today

Agentic AI Infrastructure for magnifying HUMAN capabilities.

中文介绍 面向个人的代理式AI基础设施,旨在增强人类能力而非取代。提供了构建个性化AI助手的框架与工具,适用于希望借助AI提升学习、创作或工作效率的技术用户。

santifer/career-ops

JavaScript · ★ 50,602 · 🍴 10,338 · 📈 308 stars today

AI-powered job search system built on Claude Code. 14 skill modes, Go dashboard, PDF generation, batch processing.

中文介绍 基于Claude Code的AI求职系统,内置14种技能模式,并提供Go仪表盘、PDF生成与批处理功能。适用于求职者自动化简历优化、岗位匹配、申请流程等,也可被招聘人员用于筛选候选人。

phuryn/pm-skills

★ 12,770 · 🍴 1,503 · 📈 164 stars today

PM Skills Marketplace: 100+ agentic skills, commands, and plugins — from discovery to strategy, execution, launch, and growth.

中文介绍 产品经理技能市场,提供100多个代理技能、命令和插件,覆盖从需求发现、策略制定到执行发布与增长的全流程。帮助PM通过AI驱动的方式提升工作效率和决策质量。

openai/plugins

JavaScript · ★ 2,339 · 🍴 289 · 📈 296 stars today

OpenAI Plugins

中文介绍 OpenAI官方插件仓库,为ChatGPT等产品提供第三方扩展能力。开发者可通过这些插件让AI模型访问实时信息、执行操作或集成外部服务,扩展对话式AI的应用边界。

Andyyyy64/whichllm

Python · ★ 3,528 · 🍴 204 · 📈 143 stars today

Find the local LLM that actually runs and performs best on your hardware. Ranked by real, recency-aware benchmarks, not parameter count. One command, run it instantly.

中文介绍 一键运行即可根据本地硬件实际性能,对可用的本地LLM进行基准测试并排序,结果基于真实性能而非参数数量。解决了“哪个模型在我设备上最快最好”的痛点,适合本地部署LLM的开发者。

MemPalace/mempalace

Python · ★ 54,986 · 🍴 7,160 · 📈 170 stars today

The best-benchmarked open-source AI memory system. And it's free.

中文介绍 开源AI记忆系统,自称在基准测试中表现最佳且完全免费。为AI代理提供长期记忆能力,使其能在多次交互中保留上下文和知识,适用于对话、个人助理等需要持久记忆的应用。

roboflow/supervision

Python · ★ 42,394 · 🍴 3,786 · 📈 1,288 stars today

We write your reusable computer vision tools. 💜

中文介绍 提供可复用的计算机视觉工具库,封装了目标检测、分割、追踪等常见任务的实用函数,帮助开发者快速构建CV应用而不必重复造轮子。适合计算机视觉领域的研究人员和工程师。

CopilotKit/CopilotKit

TypeScript · ★ 34,183 · 🍴 4,306 · 📈 378 stars today

The Frontend Stack for Agents & Generative UI. React, Angular, Mobile, Slack, and more. Makers of the AG-UI Protocol

中文介绍 面向AI代理和生成式UI的前端开发栈,支持React、Angular、移动端、Slack等平台,并首创AG-UI协议。旨在简化带AI助手的交互界面构建,让开发者快速集成聊天、工具调用等功能。

TapXWorld/ChinaTextbook

Roff · ★ 73,035 · 🍴 16,345 · 📈 592 stars today

所有小初高、大学PDF教材。

中文介绍 收集了中国小初高及大学各学科的PDF教材,涵盖主流版本。为学生、教师和家长提供免费的电子教材资源,方便离线阅读和查阅。

luongnv89/claude-howto

Python · ★ 35,835 · 🍴 4,350 · 📈 312 stars today

A visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value.

中文介绍 一份可视化的Claude Code学习指南,以实例驱动的方式从基础概念讲至高级代理,提供可直接复制使用的模板。帮助开发者快速上手Claude Code并构建实用Agent。

aaif-goose/goose

Rust · ★ 48,147 · 🍴 5,070 · 📈 699 stars today

an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM

中文介绍 开源、可扩展的AI代理,不仅限于代码建议,还能安装、执行、编辑和测试任务,并支持接入任意LLM。适用于开发者自动化日常流程、代码管理或系统操作。

Your UnEmbedding Matrix is Secretly a Feature Lens for Text Embeddings

👍 71

Large language models exhibit impressive zero-shot capabilities across a wide range of downstream tasks. However, they struggle to function as off-the-shelf embedding models, leading to suboptimal performance on massive text embedding benchmarks. In this paper, we identify a potential cause underlyi

中文介绍 发现LLM的UnEmbedding矩阵可充当文本嵌入的「特征透镜」,揭示其作为嵌入模型表现不佳的原因。通过分析该矩阵,提出改进方法,提升文本嵌入质量。

SoCRATES: Towards Reliable Automated Evaluation of Proactive LLM Mediation across Domains and Socio-cognitive Variations

👍 43

Evaluating LLM mediators remains challenging, as mediation unfolds as a real-time trajectory shaped by disputants' shifting emotions, intentions, and context. Existing testbeds rely on a few expert-authored domains, vary mainly strategic posture, and score every turn against every topic, introducing

中文介绍 提出SoCRATES评估框架,用于可靠自动评估LLM调解员在多领域和社会认知变化下的表现。其模拟实时调解轨迹,考虑情感与意图动态,弥补现有评估的不足。

GENEB: Why Genomic Models Are Hard to Compare

👍 42

Progress in genomic foundation models is difficult to assess due to fragmented benchmarks, incompatible evaluation protocols, and task-specific reporting. As a result, claims of superiority or generality across models are often not directly comparable. We introduce GENEB, a large-scale diagnostic be

中文介绍 引入GENEB,一个大规模诊断基准,解决基因组模型难以比较的问题。它统一碎片化基准和评估协议,提供公平对比,揭示模型在不同任务上的真实表现。

MMAE: A Massive Multitask Audio Editing Benchmark

👍 39

We introduce MMAE, a Massive Multitask Audio Editing benchmark, serving as the first comprehensive evaluation testbed designed for general-purpose instruction-based audio editing. Spurred by the shift toward intelligent creation, interactive editing has rapidly expanded from visual domains, pioneere

中文介绍 提出MMAE,首个面向通用指令式音频编辑的大规模多任务基准。覆盖多种编辑任务,推动智能音频编辑评估标准化,为模型比较提供统一平台。

SWE-Explore: Benchmarking How Coding Agents Explore Repositories

👍 24

Repository-level coding benchmarks such as SWE-bench have driven a rapid surge in the capabilities of coding agents. Yet they usually treat coding tasks as a holistic, binary prediction problem (e.g., resolved or unresolved), neglecting fine-grained agent capabilities such as repository understandin

中文介绍 提出SWE-Explore基准,细粒度评估编码代理人仓库探索能力(如理解结构、定位相关代码),弥补SWE-bench等整体二分类评估的不足。

AnchorWorld: Embodied Egocentric World Simulation with View-based Evolution Customization

👍 24

Despite being a pivotal frontier, interactive world modeling remains underexplored in terms of the versatile controllability required by practical scenarios. To bridge this gap, we present AnchorWorld, a framework that advances egocentric simulation through enhanced interaction integrity and a flexi

中文介绍 提出AnchorWorld框架,实现具身自我中心世界模拟,支持基于视角演化的自定义控制。增强交互完整性,提供灵活的可控性,推动交互世界建模实用化。

Robots Need More than VLA and World Models

👍 22

Generalist robot intelligence is often framed as a policy-scaling problem: collect more robot demonstrations, train larger Vision-Language-Action (VLA) models, and expect broader generalisation. In this position paper, we argue that this framing is incomplete. The central bottleneck is not only poli

中文介绍 观点文章指出,通用机器人智能的瓶颈不仅是策略缩放(如VLA模型),还涉及世界模型、认知结构等。提出机器人需要超越VLA和世界模型的更全面框架。

On the Geometry of On-Policy Distillation

👍 19

On-policy distillation (OPD) is increasingly used to improve large language model reasoning, but its training dynamics remain poorly understood. We characterize the trajectory of OPD updates in parameter space and compare it with supervised fine-tuning (SFT) and reinforcement learning with verifiabl

中文介绍 分析在线策略蒸馏(OPD)在参数空间中的更新轨迹,并与SFT和RL对比。揭示OPD的几何特性,解释其提升LLM推理能力的原因。

When Tools Fail: Benchmarking Dynamic Replanning and Anomaly Recovery in LLM Agents

👍 19

Existing benchmarks evaluate Tool-Integrated Reasoning (TIR) in LLMs on idealized ''happy paths'', largely overlooking real-world tool failures. We introduce ToolMaze, a benchmark for dynamic path discovery and error recovery in TIR agents. To separate systematic replanning from blind trial-and-erro

中文介绍 提出ToolMaze基准,评估LLM工具集成推理中的动态路径发现与异常恢复能力。区分系统化重规划与盲目试错,填补真实工具故障评估空白。

SubtleMemory: A Benchmark for Fine-Grained Relational Memory Discrimination in Long-Horizon AI Agents

👍 18

Persistent AI assistants, such as OpenClaw, accumulate large collections of related memories over long-term interactions. As these memories grow, they may reinforce one another, diverge across contexts, or directly conflict, making correct assistance depend on memory relations rather than isolated r

中文介绍 提出SubtleMemory基准,评估长期AI代理对细粒度关系记忆的辨别能力。测试代理在记忆冲突与上下文变化中的表现,推动持久助理的精准记忆管理。

LLM Explainability with Counterfactual Chains and Causal Graphs

👍 14

Causal graphs provide a high-level language for making mechanisms transparent. Recent work uses Large Language Models (LLMs) to recover causal graphs of external-world processes. Instead, in this paper, we use causal graphs to model LLM inference itself, providing stakeholders with a transparent vie

中文介绍 利用因果图对LLM推理过程建模,结合反事实链提供可解释性。不同于用LLM恢复外部因果图,该方法直接解释LLM内部决策机制。

Human Psychometric Questionnaires Mischaracterize LLM Behavior

👍 13

We examine whether human psychometric questionnaires can serve as reliable tools for characterizing and predicting LLM behavior in everyday user interactions. We analyze eight open-source LLMs by comparing their value and personality profiles derived from two different methods: Likert self-reports o

中文介绍 通过八种开源LLM对比Likert自我报告与行为观察,发现人类心理测量问卷会误表征LLM行为。指出应谨慎使用问卷评估LLM人格与价值观。

UnpredictaBench: A Benchmark for Evaluating Distributional Randomness in LLMs

👍 13

We introduce UnpredictaBench, an evaluation that tests the ability of large language models (LLMs) to capture true underlying distributions. As LLMs are increasingly used as substitutes for other entities (e.g., for humans in economic simulations), the tendency of many models to collapse towards a s

中文介绍 提出UnpredictaBench基准,评估LLM捕捉真实分布随机性的能力。针对模型输出分布坍塌问题,测试其在模拟场景中的分布保真度。

Watch, Remember, Reason: Human-View Video Understanding with MLLMs

👍 13

Video understanding is being rapidly transformed by multimodal large language models (MLLMs), as research moves from short clips to long, multimodal, and knowledge-intensive video scenarios. These scenarios require models to handle sparse evidence, long-range dependencies, multimodal alignment, and

中文介绍 推动MLLM从短片段向长视频、多模态知识密集型场景理解。提出「观察-记忆-推理」范式,应对稀疏证据、长程依赖和多模态对齐等挑战。

Physics in 2-Steps: Locking Motion Priors Before Visual Refinement Erases Them

👍 12

Image-to-Video diffusion models leverage input images to generate visually stunning content, yet frequently produce motion that violates physical laws. We reveal a surprising finding: a 2-step generation often exhibits better physical consistency than a 50-step output from the same model. Through sp

中文介绍 发现图像到视频扩散模型中,2步生成比50步生成物理一致性更好。原因是视觉细化阶段可能消除运动先验,提出锁定运动先验的方法以提升物理合理性。

Thinking with Imagination: Agentic Visual Spatial Reasoning with World Simulators

👍 12

While Vision-Language Models (VLMs) have shown strong visual reasoning capabilities, their spatial reasoning abilities remain largely constrained to the observed images and text-oriented chain-of-thought. They often struggle to infer unobserved layouts, maintain cross-view consistency, and reason fr

中文介绍 提出利用世界模拟器增强VLM的视觉空间推理,使其能想象未观察场景、保持跨视图一致性。结合世界模型与思维链,提升自主空间推理能力。

SIA: Self Improving AI with Harness & Weight Updates

👍 9

Humans are the bottleneck in building and improving AI. Both the models and the agents that wrap them are written, tuned, and corrected by people. The long-horizon goal of an AI that can figure out how to improve itself remains open. Two largely disjoint research lines attack this bottleneck. The ha

中文介绍 提出SIA框架,通过「束缚」与「权重更新」实现AI自我改进。结合两条独立研究线,使AI能自动调整自身模型与代理行为,减少人类干预。

ResearchClawBench: A Benchmark for End-to-End Autonomous Scientific Research

👍 5

AI coding agents are increasingly used for scientific work, but their end-to-end autonomous research capability remains difficult to verify. We present ResearchClawBench, a benchmark for evaluating autonomous scientific research across 40 tasks from 10 scientific domains. Each task is grounded in a

中文介绍 提出ResearchClawBench,评估AI代理端到端自主科学研究能力,覆盖10个领域40个任务。每个任务基于真实论文,测试完整研究流程。

SPACENUM: Revisiting Spatial Numerical Understanding in VLMs

👍 5

Vision-Language Models (VLMs) are increasingly deployed in embodied environments, where they need produce numerical outputs such as action magnitudes and spatial coordinates. Although these numbers appear meaningful, it remains unclear whether these numerical outputs are genuinely grounded in spatia

中文介绍 重新审视VLM的空间数值理解能力,发现其输出看似合理但未必真实基于空间感知。提出SPACENUM评估方法,区分真实理解与表面拟合。

Almieyar-Oryx-BloomBench: A Bilingual Multimodal Benchmark for Cognitively Informed Evaluation of Vision-Language Models

👍 4

Despite the rapid progress of Vision-Language Models (VLMs), the field lacks benchmarks that rigorously diagnose their true reasoning abilities and chart meaningful progress toward human-like multimodal intelligence. Most existing evaluations focus on piecemeal or disconnected tasks, obscuring criti

中文介绍 提出双语多模态基准Almieyar-Oryx-BloomBench,基于认知科学设计,系统诊断VLM的真实推理能力。覆盖多维度任务,推动类人多模态智能评估。

Harness Engineering: What Every AI Engineer Needs to Know in 2026

@sairahul1 · 111.8K 粉丝 · 546.4K 阅 · 536 赞 · 94 转

In February 2026, a small OpenAI team shipped 1 million lines of production code. They didn't write a single line by hand. The AI agents wrote it. The humans designed the system that made the agents

中文介绍 2026年2月,OpenAI小团队用AI代理生成100万行生产代码,人类零手写。强调「Harness Engineering」——工程师角色从写代码转变为设计驱动代理的系统,成为AI工程师必备认知。

Do AGENTS.md Files Actually Help Coding Agents?

@rasbt · 459.9K 粉丝 · 50.9K 阅 · 507 赞 · 51 转

Catching up with the agent-related research literature, one paper that definitely got my attention is "Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?." It looks

中文介绍 一篇评估AGENTS.md文件对编码代理帮助的研究论文引发关注。探讨仓库级上下文文件是否真正提升代理性能,为开发者提供是否采用的参考依据。

WTF Is a Loop? Peter Steinberger vs. Boris Cherny

@mvanhorn · 30.8K 粉丝 · 45.6K 阅 · 567 赞 · 47 转

The most repeated sentence in AI coding this week is six words long, and almost nobody saying it can define it. One tweet had the entire timeline in a chokehold this week, so I ran /last30days on the

中文介绍 AI编码圈热词「loop」几乎无人能准确定义。博主分析Peter Steinberger与Boris Cherny的争论,揭示循环在AI编程中的真实含义及工程实践误区。

17 prompts that make Hermes run while you sleep (copy-paste inside)

@Mnilax · 7.3K 粉丝 · 43.5K 阅 · 502 赞 · 42 转

In February 2026, Nous Research released Hermes Agent: an open-source, self-hosted agent that doesn't live inside an IDE and doesn't forget when the tab closes. It runs as a daemon on your own box,

中文介绍 Nous Research发布开源代理Hermes Agent,可脱离IDE作为守护进程运行。博主提供17条可复制粘贴的prompts,让代理在用户睡眠时自动工作,实现无人值守任务执行。

I Built an Agentic Harness From Scratch. That Taught Me What Agents Actually Are

@ByteMohit · 2.0K 粉丝 · 43.3K 阅 · 501 赞 · 51 转

Everyone is building with agents. Almost nobody talks about what is actually inside one. Not the model. The harness around it. I spent the last few months building one from scratch in Python every

中文介绍 博主从零用Python构建代理harness(框架),揭示代理核心不是模型而是围绕它的harness。分享构建过程中的理解:harness定义代理行为边界与决策逻辑。

Loop Engineering.

@addyosmani · 395.5K 粉丝 · 42.7K 阅 · 577 赞 · 62 转

Loop engineering is replacing yourself as the person who prompts the agent. You design the system that does it instead. A loop here can be thought of a recursive goal where you define a purpose and

中文介绍 引入「Loop Engineering」概念——设计系统代替人工手动提示代理。循环(loop)被定义为递归目标:设定目的与终止条件后,代理自主迭代执行直至完成。

RL Interview Questions 2026

@sheriyuo · 8.6K 粉丝 · 30.6K 阅 · 512 赞 · 44 转

After seeing several people receive PhD offers and then immediately land highly paid industry positions during spring recruiting, I started wondering whether going straight into industry might

中文介绍 分享2026年强化学习面试题,结合多个PhD获得者立即获得高薪工业职位的现象,探讨直接进入工业界还是读博的路径选择,为RL求职者提供参考。

Confidential submission of draft S-1 to the SEC

OpenAI confirms a confidential S-1 submission to the SEC and has not yet determined timing for further action.

中文介绍 OpenAI向美国证券交易委员会秘密提交了S-1注册声明草案,尚未确定下一步行动时间。

Built to benefit everyone: our plan

A vision for the future of AI, focusing on access, safety, and shared prosperity as OpenAI works to ensure AGI benefits everyone.

中文介绍 OpenAI发布未来AI愿景,聚焦可及性、安全性和共享繁荣,致力于确保通用人工智能(AGI)惠及所有人。

Introducing the OpenAI Economic Research Exchange

OpenAI launches the Economic Research Exchange to study AI’s impact on jobs, productivity, and the economy. Applications are now open for selected research projects.

中文介绍 OpenAI推出经济研究交流平台,旨在研究AI对就业、生产力和经济的影响,已开放研究项目申请。

[AINews] not much happened today

a quiet day of RSI.

中文介绍 AI新闻栏目报道当天相对平静,技术指标RSI显示市场平稳。

How to Stop Shipping Low-Quality RL Environments (with Examples)

Your broken harness is actively making the model worse. Here's what I keep seeing after years of eyeballing trajectories, and what you need to fix.

中文介绍 文章指出低质量的强化学习环境会降低模型性能,并提供多年观察后的修复建议与实例。

The Meta hack shows there’s more to AI security than Mythos

On June 5, 404 Media reported that attackers had been using Meta’s AI customer support agent to steal Instagram accounts. Their approach was simple: They asked the agent to link the accounts to email addresses that they controlled, and the agent complied. One attacker broke into the dormant Obama Wh

中文介绍 2026年6月5日,404 Media报道攻击者利用Meta的AI客服机器人窃取Instagram账号,通过让机器人将账户链接到攻击者控制的邮箱,暴露了AI安全风险。

not much happened today

**Anthropic's Mythos/Opus cycle** sparked mixed reactions with praise for **Claude Mythos**'s one-shot workflows and concerns over **Opus 4.8** benchmark regressions. **Opus 4.7** showed strong chemistry task performance, "making Claude a chemist." **Sakana AI** launched an **RSI Lab** focusing on r

中文介绍 Anthropic的Mythos/Opus模型周期引发不同反应:Claude Mythos的一步工作流受赞誉,Opus 4.8基准测试出现倒退;Opus 4.7在化学任务表现出色;Sakana AI推出了RSI项目。

Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs

We talk with the VendingBench authors on evaling Claudes from Haiku to Mythos, and how they build leading, and lasting, frontier evals from scratch.

中文介绍 与VendingBench作者Lukas Petersson和Axel Backlund(Andon Labs)对话,讨论评估从Haiku到Mythos系列Claude模型,以及从零构建前沿长期评估的方法。

How Endava is redesigning software delivery around AI agents

Learn how Endava is using AI agents, ChatGPT Enterprise, and Codex to accelerate software delivery, automate workflows, and build an AI-native culture across the enterprise.

中文介绍 Endava利用AI智能体、ChatGPT Enterprise和Codex重新设计软件交付流程,加速开发、自动化工作流,并构建企业AI原生文化。

市场总览

当前市场技术面呈现显著分化。美股方面(SPY、QQQ等),多数标的仍处于均线多头排列格局,但MACD普遍出现死叉,短期动能减弱,RSI多位于40-55之间,整体偏向震荡整理。加密市场受恐慌情绪主导,加密恐慌贪婪指数仅10(极度恐慌),总市值2.24T,24h微跌0.63%,BTC主导率56.1%,ETH占9%;比特币、以太坊、Solana等均处于RSI超卖区域(25-27)且空头排列,下行压力极大。中概股方面,阿里巴巴、拼多多、京东等均处于空头排列,RSI低于40,弱势明显,腾讯控股虽单日反弹4%但均线仍空头。商品外汇领域:黄金期货RSI 36.3偏弱,原油RSI 44中性,美元指数和10Y美债收益率均维持在均线上方且RSI偏强(59-64),美元指数接近52周高点,呈现上行趋势。VIX恐慌指数日内大跌12%但5日仍涨18%,技术面偏强但仍需警惕。总体而言,目前市场风险偏好较低,加密和中概承压,而美元、美债收益率延续强势。

今日关注

SPY S&P 500 ETF (SPY)
中性

RSI 14为50.8处于中性区间,价格739.22低于SMA20(746.37)但高于SMA50(715.39)和SMA200(684.43),均线多头排列维持长期上升格局,但MACD形成死叉(8.46<11.34),短期动能减弱,多空因素交织,技术面中性。

BTC-USD 比特币 (BTC-USD)
偏下行

RSI 14为25.4处于超卖区域,价格62,680.97远低于SMA20(70,969)、SMA50(75,580)和SMA200(78,355),均线空头排列,MACD柱状线为负且MACD低于信号线,空头动能主导,技术面偏下行。

META Meta (META)
偏下行

RSI 14为39.5低于50,价格585.39跌破所有均线(SMA20 611.51, SMA50 620.28, SMA200 661.64),均线空头排列,MACD死叉且柱状线-5.75,负值扩大,下跌趋势明显,技术面偏下行。

^TNX 10Y 美债收益率 (^TNX)
偏上行

RSI 14为59.1偏强,价格4.55高于所有均线(SMA20 4.51, SMA50 4.41, SMA200 4.21),均线多头排列,MACD虽略低于信号线但仍为正,整体上升趋势完好,技术面偏上行。

DX-Y.NYB 美元指数 (DX-Y.NYB)
偏上行

RSI 14为63.8接近超买,价格99.96接近52周高点,高于所有均线(SMA20 99.23, SMA50 98.91, SMA200 98.62),均线多头排列,MACD金叉(0.310>0.211),看涨动能较强,技术面偏上行。

全部资产

^VIX

VIX 恐慌指数

$18.92 -12.04%
5 日
+17.88%
距 52w 高
-46.4%
RSI(14)
54.6
趋势
多头
SMA 20 / 50 / 200
17.15 / 18.75 / 18.44
MACD / 信号
-0.218 / -0.662
MACD 金叉 (1 天前)多头排列

^TNX

10Y 美债收益率 (%)

$4.55 +0.35%
5 日
+1.72%
距 52w 高
-8.9%
RSI(14)
59.1
趋势
多头
SMA 20 / 50 / 200
4.51 / 4.41 / 4.21
MACD / 信号
0.029 / 0.035
多头排列

DX-Y.NYB

美元指数 DXY

$99.96 -0.09%
5 日
+0.75%
距 52w 高
-0.7%
RSI(14)
63.8
趋势
多头
SMA 20 / 50 / 200
99.23 / 98.91 / 98.62
MACD / 信号
0.310 / 0.211
接近 52 周高多头排列

SPY

S&P 500 ETF

$739.22 +0.23%
5 日
-2.55%
距 52w 高
-2.8%
RSI(14)
50.8
趋势
多头
SMA 20 / 50 / 200
746.37 / 715.39 / 684.43
MACD / 信号
8.461 / 11.343
接近 52 周高多头排列

QQQ

Nasdaq 100 ETF

$716.07 +1.56%
5 日
-3.59%
距 52w 高
-4.4%
RSI(14)
53.6
趋势
多头
SMA 20 / 50 / 200
722.25 / 670.65 / 622.57
MACD / 信号
15.261 / 19.609
MACD 死叉 (2 天前)多头排列

AAPL

Apple

$301.54 -1.89%
5 日
-1.56%
距 52w 高
-5.0%
RSI(14)
53.3
趋势
多头
SMA 20 / 50 / 200
304.66 / 282.21 / 265.57
MACD / 信号
7.365 / 8.989
MACD 死叉 (3 天前)多头排列

MSFT

Microsoft

$411.74 -1.18%
5 日
-10.59%
距 52w 高
-25.9%
RSI(14)
45.3
趋势
中性
SMA 20 / 50 / 200
422.41 / 409.27 / 455.91
MACD / 信号
3.851 / 5.720
MACD 死叉 (1 天前)

NVDA

Nvidia

$208.64 +1.73%
5 日
-7.01%
距 52w 高
-11.8%
RSI(14)
46.7
趋势
多头
SMA 20 / 50 / 200
218.78 / 204.20 / 188.74
MACD / 信号
1.561 / 3.719
多头排列

GOOGL

Alphabet

$363.31 -1.42%
5 日
-3.47%
距 52w 高
-11.1%
RSI(14)
43.4
趋势
多头
SMA 20 / 50 / 200
383.50 / 356.15 / 304.86
MACD / 信号
0.562 / 5.596
多头排列

TSLA

Tesla

$408.95 +4.59%
5 日
-1.67%
距 52w 高
-18.0%
RSI(14)
48.0
趋势
中性
SMA 20 / 50 / 200
424.90 / 396.03 / 414.57
MACD / 信号
3.031 / 7.492

META

Meta

$585.39 -1.28%
5 日
-2.51%
距 52w 高
-26.5%
RSI(14)
39.5
趋势
空头
SMA 20 / 50 / 200
611.51 / 620.28 / 661.64
MACD / 信号
-5.752 / -3.944
MACD 死叉 (1 天前)空头排列
加密恐慌贪婪
10
极度恐慌
加密总市值
$2.24 T
-0.63% / 24h
BTC 主导率
56.1%
ETH 9.0%
24h 成交量
$91.3 B
活跃币 17,344

BTC-USD

Bitcoin

$62,680.97 -0.88%
5 日
-2.08%
距 52w 高
-50.3%
RSI(14)
25.4
趋势
空头
SMA 20 / 50 / 200
70,969.18 / 75,580.53 / 78,355.53
MACD / 信号
-4,068.195 / -2,974.382
RSI 超卖空头排列

ETH-USD

Ethereum

$1,661.11 -1.50%
5 日
-8.31%
距 52w 高
-66.5%
RSI(14)
26.5
趋势
空头
SMA 20 / 50 / 200
1,936.21 / 2,147.76 / 2,444.12
MACD / 信号
-145.307 / -113.062
RSI 超卖空头排列

SOL-USD

Solana

$65.76 -0.83%
5 日
-8.17%
距 52w 高
-74.0%
RSI(14)
26.8
趋势
空头
SMA 20 / 50 / 200
78.07 / 83.74 / 101.70
MACD / 信号
-5.706 / -3.998
RSI 超卖空头排列

BABA

阿里巴巴 (BABA)

$120.07 -0.82%
5 日
-4.25%
距 52w 高
-37.7%
RSI(14)
36.2
趋势
空头
SMA 20 / 50 / 200
130.73 / 131.00 / 149.72
MACD / 信号
-3.045 / -1.926
空头排列

PDD

拼多多 (PDD)

$82.62 -2.88%
5 日
-5.30%
距 52w 高
-40.7%
RSI(14)
33.3
趋势
空头
SMA 20 / 50 / 200
91.67 / 96.88 / 112.28
MACD / 信号
-3.921 / -3.173
接近 52 周低空头排列

JD

京东 (JD)

$28.59 -1.00%
5 日
-1.72%
距 52w 高
-22.4%
RSI(14)
39.7
趋势
空头
SMA 20 / 50 / 200
30.61 / 30.15 / 30.37
MACD / 信号
-0.422 / -0.146
空头排列

0700.HK

腾讯控股 (0700.HK)

HK$464.40 +4.03%
5 日
-3.57%
距 52w 高
-32.0%
RSI(14)
51.7
趋势
空头
SMA 20 / 50 / 200
450.69 / 474.91 / 570.91
MACD / 信号
-5.638 / -9.352
空头排列

GC=F

黄金期货

$4,358.00 +0.51%
5 日
-2.92%
距 52w 高
-22.0%
RSI(14)
36.3
趋势
中性
SMA 20 / 50 / 200
4,508.78 / 4,620.59 / 4,408.61
MACD / 信号
-74.642 / -61.234

CL=F

WTI 原油期货

$90.89 -0.45%
5 日
-3.06%
距 52w 高
-23.9%
RSI(14)
44.0
趋势
中性
SMA 20 / 50 / 200
96.20 / 97.63 / 73.07
MACD / 信号
-1.846 / -1.303

USDCNY=X

美元 / 人民币

¥6.78 +0.20%
5 日
+0.21%
距 52w 高
-6.0%
RSI(14)
43.2
趋势
空头
SMA 20 / 50 / 200
6.79 / 6.81 / 6.97
MACD / 信号
-0.014 / -0.014
MACD 金叉 (今天)接近 52 周低空头排列
风险提示

过去走势不代表未来表现,本报告仅供技术指标解读参考,不构成任何投资建议。

Jacinta Allan says Pauline Hanson ‘chooses to barrack for bullies’ over ‘ditch the witch’ billboard

Victorian premier says she will always call out ‘misogynist’ views after One Nation leader says ‘if the shoe fits’ Jacinta Allan has dismissed leadership speculation and says she will continue to call out “sexist, misogynist, hateful” commentary directed at her, despite One Nation leader Pauline Han

中文摘要 维多利亚州州长杰辛塔·艾伦否认领导权猜测,指责保琳·汉森“选择为恶霸助威”,批评其“厌女”言论。艾伦表示将继续反对性别歧视和仇恨言论。

Australia news live: Victorian premier Jacinta Allan says Pauline Hanson ‘chooses to barrack for the bullies’; ASX slumps amid tech selloff

Allan rejects leadership speculation and says she is ‘all in’. Follow live Get our breaking news email, free app or daily news podcast ‘If Australian datacentres are going to power the AI revolution, we deserve a fair return’ – David Pocock Independent senator David Pocock has published an opinion p

中文摘要 澳大利亚新闻:维州州长艾伦否认领导权猜测,称“全力以赴”;ASX因科技股抛售下跌。独立参议员波科克呼吁对人工智能数据中心获取公平回报。

Live Updates: Israel Halts Iran Strikes After Trump Claims Progress Toward Nuclear Talks, Officials Say

Prime Minister Benjamin Netanyahu said that Israel’s “fire is on hold” after a phone call with President Trump. Iran also said it would cease its attacks but, like Israel, warned it was ready to resume.

中文摘要 以色列总理内塔尼亚胡表示“火力暂停”,此前与特朗普通电话。伊朗也宣布停止攻击,但双方均警告随时可能恢复。

UN questions legality of Israeli forced evacuation orders in Lebanon

A UN spokesman says forced evacuation orders issued across southern and eastern Lebanon are nearly impossible to follow

中文摘要 联合国发言人表示,以色列在黎巴嫩南部和东部发布的强制撤离令几乎无法执行,质疑其合法性。

Barnaby Joyce claims people who ‘look like’ recent arrivals dominate auctions. Who is he talking about?

Statistics show fewer than 1% of homes are bought by foreigners, as race discrimination commissioner warns against blaming migrants for housing woes Follow our Australia news live blog for latest updates Get our breaking news email, free app or daily news podcast One Nation’s Barnaby Joyce has claim

中文摘要 巴纳比·乔伊斯声称看起来像新移民的人主导拍卖。数据显示外国人购房比例不足1%,种族歧视专员警告不要将住房问题归咎于移民。

French star Patrick Bruel held by police investigating new sexual assault allegations

Two new rape complaints have been filed against the 67-year-old singer and actor, who denies the claims French singer and actor Patrick Bruel, facing sexual assault allegations from multiple women, was taken into police custody on Monday, as two new rape complaints were filed against him. The 67-yea

中文摘要 法国歌手兼演员帕特里克·布鲁尔因新的强奸指控被警方拘留,已有两名女性提出新的强奸投诉。布鲁尔否认指控。

ICC prosecutor suspended pending vote on sexual misconduct claims

Karim Khan rejects suspension as ICC governing body refers him to disciplinary proceedings before member states.

中文摘要 国际刑事法院检察官卡里姆·汗因性行为不端指控被暂停职务,法院理事机构将其提交纪律程序,汗拒绝接受暂停。

Israel continues strikes on Lebanon despite halting attacks on Iran

Israeli PM Benjamin Netanyahu acknowledged a halt in fighting with Iran but will continue operations against Hezbollah.

中文摘要 以色列总理内塔尼亚胡承认对伊朗停止攻击,但将继续对黎巴嫩真主党的军事行动。

Peru election result close as vote counting continues

The race between right-wing Keiko Fujimori and left-wing Roberto Sánchez has been dominated by concerns over crime and political instability.

中文摘要 秘鲁总统选举计票胶着,右翼候选人藤森庆子与左翼候选人罗伯托·桑切斯竞争激烈,犯罪和政治不稳定成为主要议题。

Iran war live: Trump warns Netanyahu as Israel, Tehran halt fighting

Israeli strikes have killed 3,637 people in Lebanon since March, with 11,188 wounded, the Health Ministry says.

中文摘要 特朗普警告内塔尼亚胡,以色列与伊朗停止战斗。黎巴嫩卫生部称,自3月以来以色列袭击已致3637人死亡,11188人受伤。

Trump nominates Todd Blanche as attorney general, setting up Senate fight

Blanche, who currently serves as acting attorney general, has faced controversy over the Epstein files and January 6.

中文摘要 特朗普提名托德·布兰奇为司法部长,可能引发参议院斗争。布兰奇现任代理司法部长,曾卷入爱泼斯坦文件和1月6日事件争议。

GSK in talks to buy cancer biotech Nuvalent for more than $9bn

Deal would rank as one of UK drugmaker’s biggest as biotech M&A activity accelerates this year

中文摘要 英国药企葛兰素史克(GSK)正在谈判收购癌症生物技术公司Nuvalent,交易金额超过90亿美元,将成为GSK最大收购之一,生物技术并购活动加速。

Maybank Handles $4.9 Billion Deals for Singapore, Malaysia Hub

Malayan Banking Bhd said the Johor-Singapore special economic zone will drive further expansion, with Malaysia’s biggest lender having facilitated about 20 billion ringgit ($4.9 billion) in financing and investments in the area in the past couple of years.

中文摘要 马来亚银行(Maybank)表示,柔佛-新加坡经济特区将推动进一步扩张,该行过去两年为这一地区提供了约200亿林吉特(49亿美元)的融资和投资。

Citadel Securities Expands India Trading and Engineering Teams

Citadel Securities LLC, the market-making firm founded by billionaire Ken Griffin, is hiring more engineers and traders in India, as it seeks to expand in one of the world’s fastest-growing financial markets.

中文摘要 亿万富翁肯·格里芬创立的做市商Citadel Securities正在印度招聘更多工程师和交易员,以扩展其在全球增长最快的金融市场之一的业务。

GSK in Talks to Buy Nuvalent for More Than $9 Billion: FT

GSK Plc is in discussions to purchase cancer treatment maker Nuvalent Inc. for between $9 billion and $10 billion, the Financial Times reported, citing unidentified people familiar with the matter.

中文摘要 据英国《金融时报》报道,英国药企GSK正在讨论以90亿至100亿美元收购癌症治疗公司Nuvalent。

Bain Capital Makes Non-Binding Bid for Australia’s oOh!Media

Bain Capital has joined the pack of bidders for oOh!Media Group, the Australian outdoor advertising company said.

中文摘要 贝恩资本加入了澳大利亚户外广告公司oOh!Media的竞购阵营,提交了非约束性报价。

Xi Receives Lavish Welcome From Kim in North Korea

President Xi Jinping arrived in Pyongyang to an elaborate welcome on his first visit to the reclusive nation in seven years In a meeting with Kim Jong Un, Xi vowed to deepen China's ties with North Korea. (Source: Bloomberg)

中文摘要 中国国家主席习近平抵达平壤,受到朝鲜领导人金正恩的隆重欢迎,这是习近平七年来首次访问朝鲜。习近平承诺深化中朝关系。

Apple unveils ‘Siri AI’ in challenge to rival chatbots

Silicon Valley giant promises user privacy as it makes long-delayed overhaul of voice assistant

中文摘要 苹果公司发布名为「Siri AI」的语音助手重大升级,挑战竞争对手聊天机器人,并承诺保护用户隐私。

FT Financial Literacy and Inclusion Campaign

The FT invites readers to join our campaign to promote financial literacy in the UK and around the world

中文摘要 英国《金融时报》邀请读者加入其在英国及全球推广金融素养的公益行动。

Morgan Stanley Sees LNG Upside Risks as Asian Demand Picks Up

Liquefied natural gas prices are set to climb to levels not seen in more than three years as hotter weather in Asia and restocking needs in Europe boost demand, according to Morgan Stanley.

中文摘要 摩根士丹利认为,随着亚洲炎热天气和欧洲补库需求推动,液化天然气(LNG)价格将升至三年多来最高水平。

Apollo and Blackstone raise $35bn in chip financing deal for Anthropic

Transaction is one of the largest private credit fundraisings, fuelling the Claude maker’s AI growth plans

中文摘要 Apollo和黑石集团为Anthropic筹集350亿美元芯片融资交易,这是最大规模私人信贷融资之一,支持Claude开发商的AI扩张。

SpaceX's stock market blast-off could be Musk's biggest gamble yet

SpaceX is preparing for a stock market debut that could transform the company, the wider market and Elon Musk's fortune.

中文摘要 SpaceX正准备首次公开募股(IPO),这可能彻底改变公司、市场和埃隆·马斯克的财富。

一种更高效的蹬公益站的方式【续】

先放一张最终配置完成后的效果图: 看到 @Gnn 佬的帖子后,发现很多佬友在评论区问,CC Switch 里的用量查询是怎么配置的,刚刚研究了下搞懂了,顺手整理出来分享一下。 首先可以先根据 @Gnn 佬的帖子配置【自动故障转移】;配置完成后,CC Switch 就可以自动切换到当前更合适的公益站节点;帖子各位佬友可以去看看;不过这个配不配都无所谓,本帖讲的主要是用量查询的配置。 接下来配置用量统计和展示 在 CC Switch 里点击【配置用量查询】: 进去后只需要填三个东西:请求地址、用户 ID、访问令牌。就能在 CC Switch 列表里看到每个公益站的已使用额度和剩余额度了。 具体对应

关于最近几天 君の的公益 签到都是520的一些解释

其实我真的很忙,我也不知道我在忙什么,反正“降低签到额度” 这件事情的优先级很低,我就一直没弄,那个额度其实意义不大,我也不建议大家囤着不用,毕竟用出去的tokens,才是实在的,偶尔号池波动造成的卡顿其实是难免的(例如空回)。 因为并发和调用量已经达到了一个惊人的地步 就这样,我把额度改回52刀了,期待下一次狂欢吧,或许是端午节呢? 祝大家早安,午安,晚安~ 对了,大家遇到问题回帖询问即可,不需要单独开贴占用社区资源; 感谢很多佬友的无偿帮助,没有他们,就没有 君の的公益 151 个帖子 - 147 位参与者 阅读完整话题

AI一次性跑新高考 I 卷数学究竟能拿多少分?

从 【已公布部分结果,继续测其他模型~】佬们觉得哪个AI高考数学肯定能考满分? 以及 新高考数学一卷出炉,测测哪些 AI 有实力 继续讨论 本次测试为一次性全部发送,看模型能答多少分 叠甲: 问 1: 答 1: 问 2: 答 2: 会补上的 @Nobody_233 佬帮忙测试一次(官网 max thinking) 问 3: 答 3: 问 4: 答 4: 模型环境 GPT-5.2 Pro (官网 Extended Pro);GPT-5.5 / GPT-5.4 / GPT-5.2 Thinking(推理强度:Extra High): Gemini Deep Think: Gemini 3.1 Pr

【开源推广】作为一名在读博士生,我在日常是如何与 AI 协作的?——ai-collab-playbook(26.6.8版)

作为一名在读博士生,我在日常是如何与 AI 协作的?——ai-collab-playbook > 公开版本 / Public edition: 2026-06-08 cnfjlhj/ai-collab-playbook github版本~ -— 前言 我是一名人工智能方向的在读博士生,大概在 ChatGPT 出来以后还是 GPT-3.5 的时候就比较重度使用 AI 以及 AI 工具了。几年下来,AI 已经渗透到我工作和学习很多环节,有一些心得想分享一下~ 当同事,不当工具(我认为至少未来几年,应该是人机协作的时代) 现在回头看,我觉得过去很多所谓“会用电脑”,其实有相当一部分是在给机器当翻译。

OpenCodex 2.0版本!新架构!随时随地AI Coding!

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 好久不见,这段时间实在太忙,而且一直在优化和测试新的架构,现在终于能给大家发最新版本了! 新版本不再需要繁琐的实现IPC,最大程度兼容Codex!Codex更新也不会导致功能大面积失效!还提高了数据加载速度和完全兼容M

在中转妖魔鬼怪都存在的时代,你们可能需要隐私过滤,于是我们开源了

privacy-filter 纯 Go 实现,零依赖,单个二进制文件 支持三种接入方式:import 包、HTTP 接口、gRPC 两层检测机制:结构化 PII 识别 + Gitleaks 全套 222 条规则 之前在 PackyCode 上线隐私过滤功能后,收到不少反馈和需求,所以直接开源了。 欢迎大家 star、试用、提 issue、PR 项目地址:https://github.com/packyme/privacy-filter 为了省事我直接照搬推特发的内容了 这个玩意儿已经在packyapi 上面内置了,有需要的可以自取 20 个帖子 - 14 位参与者 阅读完整话题

又来修一下CC

问题是这样的 从 2.1.140+ (大约 144)开始 发现 Bash权限审批的弹窗搁那不弹出来还卡住了 现象是 ⏺ Bash(tar tzf /Users/haleclipse/WorkSpace/Node/ClaudeCodeRev/rc-server/dist/rc-server-0.1.0.tgz | wc -l; tar tzf /Users/haleclipse/WorkSpace/Node/ClaudeCodeRev/rc-ser…) ⎿ Waiting… ✶ Frosting… (5m 54s · ↓ 2.0k tokens) 嗯 就硬 Waiting… 不走了 头回发现的时

兄弟们,感觉被人作局了?

有人跟我说,5年 GitHub 账号没多少人的,能进的基本上之前也都进过了。我信了,开了这个渠道,结果 6 个小时整了差不多 5k 人进来?GitHub oauth API 都搞出 Rate limit 了? 这玩意据说是 5000次/1H 啊,这也能爆掉?幸亏 5.1 的服务器扩容给力。 这指定是被作局了,这另一条腿我根本都不敢出了。 当然,还是要欢迎新来的佬友,请务必仔细看一看我们的社区准则,希望你们在这里玩得愉快: https://linux.do/guidelines 447 个帖子 - 419 位参与者 阅读完整话题

友善提醒下刚来的新人们

发本帖的目的是,刚刚涌入了一些新的佬友们,似乎对咱们社区的一切不很熟悉。 因此,在我刚刚水帖的几分钟内,已经看到10个以上的佬友有违规行为。 特发此帖,予以提醒。 (点击了解更多详细信息) [!danger] 这些是重中之重,你必须阅读!! 首先,社区很欢迎你的加入! 始皇的表态: 兄弟们,感觉被人作局了? 当然,还是要欢迎新来的佬友,请务必仔细看一看我们的社区准则,希望你们在这里玩得愉快: https://linux.do/guidelines 但是,我们希望你能够认真地阅读准则,这是作为一名成员的基本素养。 然后,请注意:无数佬友踩坑的规定:非必要不抽奖,非必要不抽奖,非必要不抽奖。重要的