2026-07-21（内容获取于 07/21 05:03）

AI’s most important protocol is getting a little bit easier to use

TechCrunch · 07/21 04:50

The Model Context Protocol (MCP) is one of the basic building blocks of AI interoperability, giving AI models a secure way to access external data sources and services. It’s the plumbing that lets a chatbot reach into your calendar, your database, or your internal tools, instead of engineers buildin

中文介绍 Model Context Protocol (MCP) 作为 AI 互操作性的基础构建，正在变得更易使用。该协议为 AI 模型提供了安全访问外部数据源和服务的途径，使聊天机器人等应用能够安全地调用日历、数据库或内部工具。

The Space Force is now seeking to buy up to $30 billion in rocket launches

Ars Technica · 07/21 04:30

The Trump administration is asking the Space Force to do a lot. This will require more launches.

中文介绍 美国太空军正寻求购买价值高达 300 亿美元的火箭发射服务，以支持特朗普政府赋予其的扩展任务需求。此举旨在满足太空军未来日益增长的大量发射任务需求。

Get Borderlands 3, Risk of Rain 2 and 13 other great PC games for $15

The Verge · 07/21 04:08

Borderlands 3 is among the 15 games in the 2K Megahits 2026 Bundle. | 2K Games The aptly-named “2K Megahits 2026 Bundle” from Humble includes 15 Steam games for $15. It’s a smattering of the publisher 2K’s biggest games of the 2010s, including BioShock Infinite, Risk of Rain 2, Borderlands 3, and XC

中文介绍 Humble Bundle 平台推出「2K Megahits 2026」游戏合集，以 15 美元的价格提供 15 款 Steam 游戏。该合集包含 2K 公司在 2010 年代发行的多款热门作品，如《无主之地 3》、《雨中冒险 2》和《生化奇兵：无限》。

The PlayStation replica ornament is an homage to a great, yet fragile console

The Verge · 07/21 04:06

The ornament requires three 1.5-volt coin batteries, which are included in the box. | Photo: Cameron Faulkner/The Verge You probably know the signature PlayStation boot sound. Did you know that it's technically a multi-part chime? There's the synthy section where "Sony Computer Entertainment" shows

中文介绍 Hallmark 推出 PlayStation 复制品装饰摆件，这款致敬经典游戏机的产品可播放标志性的 PlayStation 开机音效。该摆件需三节 1.5 伏纽扣电池供电，电池已包含在包装内。

X relaunches a rebuilt Android app after year-long effort

TechCrunch · 07/21 03:37

X says the rebuilt version of its Android app is now available globally.

中文介绍 社交媒体平台 X (前身为 Twitter) 宣布，经过长达一年的重建工作，其 Android 应用程序的全新版本现已在全球范围内重新上线。

OpenAI is scared of open-weight models. Should the US be?

TechCrunch · 07/21 03:33

Talk of banning Chinese-made open-weight LLMs reveals the challenge of turning AI into a business.

中文介绍 OpenAI 对开源大模型表达担忧，而关于禁止中国制造的开源大语言模型的讨论，则揭示了将人工智能技术商业化所面临的挑战。这引发了美国是否也应警惕此类模型的疑问。

AT&T loses key ruling in bid to stop offering basic phone service in California

Ars Technica · 07/21 03:20

AT&T suffers setback but will keep asking court and FCC to preempt state rules.

中文介绍 AT&T 在其寻求停止在加州提供基础电话服务的尝试中遭遇关键裁决失利。尽管受挫，该公司表示将继续向法院和联邦通信委员会 (FCC) 提出申请，以期超越州级法规。

Ford’s $30,000 electric truck: all the news about the company’s big EV re-do

The Verge · 07/21 03:18

The end of the Ford F-150 Lightning was also the start of a new era for the automaker. After failing to capture the market with the electric version of its popular F-Series truck, Ford is going back to the drawing board with its Universal Electric Vehicle Platform. The first offering will be a $30,0

中文介绍 福特汽车正重新规划其电动汽车战略，计划推出一款售价 30,000 美元的电动皮卡。此前，F-150 Lightning 电动版未能成功占领市场，促使福特转向全新的通用电动汽车平台进行开发。

Natural raises $30M to reinvent payments for AI agents — and take on Stripe

TechCrunch · 07/21 03:11

The one-year-old startup aims to reinvent financial architecture for autonomous AI transactions.

中文介绍 成立一年的初创公司 Natural 成功筹集 3000 万美元资金，旨在彻底改革 AI 代理的支付系统，并挑战 Stripe 等现有支付巨头。该公司致力于为自主 AI 交易重塑金融架构。

Google's "Frozen v2" chip reportedly bakes Gemini's architecture directly into silicon for efficiency gains

The Decoder · 07/21 02:08

Google is developing "Frozen v2," a server chip that bakes the Gemini architecture directly into hardware. According to internal sources, it could be 6 to 10 times more efficient than current TPUs. Scheduled for 2028, the chip would drastically cut Google's AI inference costs and could give the comp

中文介绍 谷歌正在开发一款代号为「Frozen v2」的服务器芯片，据称其将 Gemini 架构直接融入硬件。内部消息显示，该芯片预计比现有 TPU 效率高出 6 到 10 倍，计划于 2028 年推出，有望大幅降低谷歌的 AI 推理成本。

RIP bargain bin: The price impact of Sony's disc-free PlayStation plan

Ars Technica · 07/21 02:06

Used discs are often cheaper than even deep digital discounts.

中文介绍 索尼推出的无光驱版 PlayStation 计划对游戏价格产生影响，分析显示实体游戏的重要性依然存在。与数字版深度折扣相比，二手光盘通常更为便宜，体现了实体游戏在价格方面的优势。

China’s AI models have Trump’s AI world at war with itself

MIT Tech Review AI · 07/21 02:00

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Over the weekend, several current and former advisors to President Donald Trump on AI publicly lobbed insults at the country’s leading AI companies. David Sacks,

中文介绍 针对中国 AI 模型，美国前总统唐纳德·特朗普的 AI 顾问团队内部出现分歧，周末多名现任及前任顾问公开抨击美国领先的 AI 公司。这凸显了美国在 AI 政策和战略制定方面的复杂矛盾。

Judge pauses $110B Paramount-Warner Bros. merger

TechCrunch · 07/21 01:58

The lawsuit from the states alleges that the deal would harm movie theaters, basic cable distributors, and audiences.

中文介绍 一名法官已暂停价值 1100 亿美元的派拉蒙 (Paramount) 与华纳兄弟 (Warner Bros.) 合并案。多州提起的诉讼指控称，这笔交易将损害电影院、有线电视基础分销商以及广大观众的利益。

Judge pauses Paramount’s attempt to buy Warner Bros. Discovery

The Verge · 07/21 01:37

A judge partially granted the request from a dozen state attorneys general to temporarily place the $110 billion merger of Paramount and Warner Bros. Discovery on hold, as reported by Variety and Reuters. US District Judge Araceli Martínez-Olguín said that based on the new company's market share, "t

中文介绍 美国地区法官 Araceli Martínez-Olguín 批准了十余个州总检察长提出的部分请求，暂时搁置了派拉蒙 (Paramount) 与华纳兄弟探索 (Warner Bros. Discovery) 之间价值 1100 亿美元的合并案。

District 9 director Neill Blomkamp releases first short film made entirely with AI video generation

The Decoder · 07/21 01:32

"District 9" director Neill Blomkamp has released "Nightborne," a 13-minute sci-fi horror short generated entirely with the Seedance 2.0 video model. Blomkamp directed frame by frame through text prompts and founded a new AI film studio, Barley Studios, to produce a full-length feature next. The art

中文介绍 电影《第九区》导演尼尔·布洛姆坎普发布了其首部完全由 AI 视频生成模型 Seedance 2.0 制作的短片《夜生》(Nightborne)，这是一部 13 分钟的科幻恐怖作品。布洛姆坎普通过文本提示逐帧指导，并成立了新 AI 电影工作室 Barley Studios。

tirth8205/code-review-graph

Python · ★ 22,945 · 🍴 2,261 · 📈 1,876 stars today

Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.

中文介绍 Code Review Graph是一个本地优先的代码智能图谱工具，专为MCP（多组件项目）和CLI环境设计。它构建了一个持久化的代码库映射，旨在优化AI编码工具的上下文理解，确保AI在代码审查和处理大型仓库时仅读取关键信息。通过大幅减少AI上下文长度，提升了代码理解效率和审查质量。

1jehuang/jcode

Rust · ★ 9,546 · 🍴 1,078 · 📈 612 stars today

The most intelligent agent harness for code

中文介绍 Jcode 是一个专为编码 Agent 设计的开发与评估框架（Coding Agent Harness）。该项目旨在为研究人员和开发者提供一套工具，用于构建、测试和比较 AI 编码助手的性能与能力。它能够帮助评估 AI 模型在代码生成、错误修复、重构以及其他编程任务中的表现。通过 Jcode，开发者可以更有效地迭代和优化 AI 驱动的编程工具，从而推动自动化编程和智能辅助编码技术的发展。

diegosouzapw/OmniRoute

TypeScript · ★ 21,550 · 🍴 2,940 · 📈 1,300 stars today

Never stop coding. Free MIT AI gateway: one endpoint, 268+ providers (50+ free), 500+ models — Claude, GPT, Gemini, Kimi K3, GLM, DeepSeek. Works with Claude Code, Codex, Cursor, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, multimodal, Desktop/PWA

中文介绍 OmniRoute 提供一个免费的 AI 网关，通过单一 API 端点集成超过231个 AI 提供商（其中50余个免费），旨在解决多模型调用的复杂性。它支持将 Claude Code、Codex、Cursor、Cline、Copilot 等编码助手连接至免费的 Claude、GPT、Gemini 等主流大模型，大幅降低开发成本。项目还采用 RTK+Caveman 堆叠压缩技术，可节省 15-95% 的数据传输开销，适合开发者统一管理 AI 服务、优化性能并利用免费资源进行高效开发。

rohitg00/ai-engineering-from-scratch

Python · ★ 40,449 · 🍴 6,716 · 📈 846 stars today

Learn it. Build it. Ship it for others.

中文介绍 ai-engineering-from-scratch 项目提供了一个从零开始学习 AI 工程的全面指南，涵盖 AI 系统从概念、开发、构建到部署的整个生命周期。旨在帮助开发者和工程师掌握构建、测试并交付可生产级 AI 应用所需的实践技能和最佳实践。它通过结构化的学习路径，适合希望系统性学习 AI 工程并提升实战能力的个人。

msitarzewski/agency-agents

Shell · ★ 134,628 · 🍴 21,956 · 📈 744 stars today

A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.

中文介绍 agency-agents 项目提供了一个完整的 AI 代理机构框架，通过部署多个具有特定人格和专业技能的 AI Agent，协同完成复杂任务。这些 Agent 涵盖前端开发、社区运营、创意生成和事实核查等多样化角色，旨在模拟人类团队协作，实现自动化决策与执行。它适用于需要集成多种 AI 能力、自动化跨领域工作流的开发者和企业。

kvcache-ai/ktransformers

Python · ★ 18,711 · 🍴 1,463 · 📈 448 stars today

A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations

中文介绍 ktransformers 是一个灵活的框架，专注于 LLM (大语言模型) 推理和微调的异构优化实践。该项目旨在帮助开发者和研究人员探索并应用多种优化技术，以提升 LLM 在不同硬件环境下的性能和效率。它简化了复杂的优化流程，使用户能够方便地试验和比较不同的优化策略，例如 KV cache 优化、量化、剪枝等，从而在部署和训练 LLM 时实现资源的高效利用和更快的响应速度。

jamiepine/voicebox

TypeScript · ★ 44,083 · 🍴 5,356 · 📈 839 stars today

The open-source AI voice studio. Clone, dictate, create.

中文介绍 Voicebox 是一个开源的 AI 语音工作室，为用户提供强大的语音克隆、文本转语音及内容创作能力。它允许用户轻松克隆特定人声，将文本转化为逼真且情感丰富的语音，并基于此生成全新的音频内容。该项目旨在降低高质量语音合成技术的门槛，赋能内容创作者、开发者以及任何对个性化语音应用有需求的用户，广泛应用于播客制作、有声读物、虚拟助手或多媒体内容配音等场景。

topoteretes/cognee

Python · ★ 28,724 · 🍴 2,740 · 📈 249 stars today

Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.

中文介绍 Cognee 是一个开源的 AI 记忆平台，旨在为 AI Agent 提供跨会话的持久性长期记忆能力。通过其自托管的知识图谱引擎，Agent 能够存储和检索信息，从而在不同会话中保持上下文和学习经验，解决传统 Agent 记忆短暂的问题。这使得开发者可以构建更智能、更具连贯性的 AI Agent，适用于需要长期记忆和复杂推理的自动化任务。

Robbyant/lingbot-map

Python · ★ 14,140 · 🍴 1,479 · 📈 554 stars today

A feed-forward 3D foundation model for reconstructing scenes from streaming data

中文介绍 `lingbot-map` 是一个面向机器人和实时应用设计的 3D 基础模型，它利用前馈神经网络技术，能从连续的流式数据中高效重建三维场景。该项目旨在解决动态环境中实时感知和高精度地图构建的挑战，尤其适用于需要即时理解周围环境的机器人、自动驾驶系统或增强现实（AR）设备。通过提供一个强大的 3D 感知框架，它能够帮助这些系统实现更智能的导航、避障和交互能力，为未来的智能体和沉浸式体验提供核心空间感知能力。

every-app/open-seo

TypeScript · ★ 5,746 · 🍴 638 · 📈 983 stars today

Open source alternative to Semrush and Ahrefs

中文介绍 open-seo 旨在成为 Semrush 和 Ahrefs 等商业 SEO 工具的开源替代品。它提供了一套免费且透明的解决方案，帮助网站管理员、营销人员和内容创作者进行关键词研究、竞争对手分析、网站审计及排名跟踪，从而优化搜索引擎表现。对于预算有限但又需要强大 SEO 功能的用户来说，open-seo 提供了一个极具吸引力的开源选择，助力网站提升自然搜索流量。

MoonshotAI/kimi-cli

Python · ★ 10,193 · 🍴 1,216 · 📈 405 stars today

Kimi Code CLI is your next CLI agent.

中文介绍 Kimi Code CLI 是由 MoonshotAI 推出的一个命令行接口 (CLI) 工具，旨在作为开发者的“下一代 CLI 代理”。它将 AI 的能力直接带入终端，允许开发者通过命令行与 AI 进行交互，执行代码辅助、信息查询或自动化编程任务。Kimi-cli 适用于希望在命令行环境中高效利用 AI 能力、提升开发效率的程序员和工程师。

AstrBotDevs/AstrBot

Python · ★ 37,038 · 🍴 2,574 · 📈 330 stars today

AI Agent Assistant & development framework that integrates lots of IM platforms, LLMs, plugins and AI feature, and can be your openclaw alternative. ✨

中文介绍 AstrBot 是一个多功能的 AI Agent 助手和开发框架，旨在简化跨平台智能助手的构建与部署。它深度集成了多种即时通讯（IM）平台、主流大语言模型（LLM）以及丰富的插件和 AI 功能。该项目提供了强大的扩展性，允许开发者根据需求定制 Agent 的行为和能力，并可作为 OpenClaw 等工具的开源替代方案。AstrBot 适用于需要构建在不同社交和办公平台上线，并能灵活响应用户请求的智能 AI 助手的开发者。

PrefectHQ/fastmcp

Python · ★ 26,490 · 🍴 2,160 · 📈 77 stars today

🚀 The fast, Pythonic way to build MCP servers and clients.

中文介绍 fastmcp 提供了一个快速且符合 Python 习惯的方式来构建 Minecraft 协议（MCP）服务器和客户端。它简化了复杂的 Minecraft 协议交互开发，让开发者能更容易地用 Python 创建自定义游戏服务器、实现自动化客户端（如机器人）或开发服务器管理工具。该项目旨在降低 Minecraft 生态系统工具和应用的开发门槛，特别适合希望构建自定义 Minecraft 体验、自动化游戏任务或进行协议级调试的开发者使用。

KnockOutEZ/wigolo

TypeScript · ★ 2,454 · 🍴 143 · 📈 695 stars today

The go-to web for your AI coding agent — local-first search, fetch, crawl & research over MCP. No API keys, no cloud, $0/query. Public beta.

中文介绍 wigolo 为 AI 编程助手提供一个本地优先的信息检索与研究平台。它允许 AI 代理在无需 API 密钥、不依赖云服务且零成本的情况下，对本地代码项目 (MCP) 进行搜索、抓取和研究。这为开发者提供了一个私密、高效且完全离线的环境，用于增强 AI 编码助手的知识库，同时保护代码隐私并大幅降低运行成本。

tokio-rs/topcoat

Rust · ★ 1,463 · 🍴 39 · 📈 374 stars today

A batteries-included framework for building web apps

中文介绍 topcoat 是一个为构建 Web 应用设计的“功能完备”（batteries-included）的 Rust 框架。它基于 Tokio 异步运行时，旨在提供一站式解决方案，简化 Rust 后端开发中的常见任务，如路由、中间件、数据库集成等。该框架帮助开发者快速启动并构建高性能、并发的 Web 服务和 API，减少了从零开始搭建项目所需的时间和精力。尤其适合寻求生产级、Rust 生态内成熟 Web 开发体验的工程师。

iptv-org/iptv

TypeScript · ★ 133,722 · 🍴 7,558 · 📈 144 stars today

Collection of publicly available IPTV channels from all over the world

中文介绍 iptv-org/iptv 汇集了全球范围内公开可用的 IPTV 频道资源。它提供了一个庞大且持续更新的M3U播放列表集合，使用户能够方便地访问来自不同国家和地区的免费电视频道。这个项目旨在简化查找和观看全球网络电视内容的过程，避免用户自行搜索的繁琐。它广泛适用于希望通过 VLC、Kodi 等媒体播放器观看国际频道的普通用户，以及开发自定义 IPTV 客户端或媒体中心的开发者。

oblien/openship

TypeScript · ★ 4,618 · 🍴 318 · 📈 1,719 stars today

Self-hosted deployment platform

中文介绍 openship 是一个开源的自托管部署平台，旨在为开发者和运维团队提供一套完整的应用发布和管理解决方案。用户可以在自己的服务器或云基础设施上部署该平台，从而拥有对应用部署流程和数据的完全控制。它通常集成了版本控制、持续集成/持续部署（CI/CD）、环境管理、扩展和监控等功能，帮助团队实现自动化、高效的应用上线。适用于需要内部部署、注重数据主权或寻求替代商业部署服务的企业和个人。

microsoft/Ontology-Playground

TypeScript · ★ 1,682 · 🍴 227 · 📈 487 stars today

Free, open-source web app for learning about ontologies and Microsoft Fabric IQ. Explore a catalogue of pre-built ontologies, design your own visually, export as RDF/XML, and share interactive diagrams. Zero backend, fully static.

中文介绍 Ontology-Playground 是微软推出的一款免费开源 Web 应用，专为学习本体论（ontologies）和 Microsoft Fabric IQ 设计。它提供了一个互动平台，用户可以探索预构建的本体目录，通过直观的图形界面设计自己的本体结构，并将其导出为标准的 RDF/XML 格式。此外，用户还能分享生成的交互式本体图。该工具旨在降低本体论的学习门槛，并为数据架构师、知识工程师及对语义数据建模和知识图谱感兴趣的开发者提供一个实用的设计与可视化环境。

handy-computer/transcribe.cpp

C++ · ★ 1,196 · 🍴 29 · 📈 401 stars today

ggml speech-to-text inference for 16+ model families

中文介绍 transcribe.cpp 是一个基于 ggml 库实现的 C++ 项目，专注于为 16 种以上模型家族提供高效的语音转文本（Speech-to-Text, STT）推理能力。它利用 ggml 的优化特性，使得 STT 推理能够在 CPU 上高效运行，降低了对专用硬件的需求。该项目解决了在本地设备、边缘计算或对隐私有严格要求的场景下进行离线语音识别的需求。开发者可以利用它将 STT 功能集成到桌面应用、嵌入式系统或任何需要低延迟、本地化语音处理的解决方案中。

moonshine-ai/moonshine

C++ · ★ 9,742 · 🍴 515 · 📈 264 stars today

Very low latency speech to text, intent recognition, and text to speech, for building voice agents and interfaces

中文介绍 moonshine 是一个专为构建语音代理和交互界面而设计的人工智能项目。它提供了一整套超低延迟的语音转文本（STT）、意图识别和文本转语音（TTS）功能。该项目旨在解决在实时对话系统中，确保自然、流畅人机交互所需的性能和响应速度挑战。开发者可以利用 moonshine 快速集成端到端语音能力，适用于开发智能语音助手、客服机器人、互动式语音响应系统或任何需要高效率语音交互的应用程序。

You just hired a million bad employees.

@gsivulka · 10.2K 粉丝 · 591.2K 阅 · 512 赞 · 99 转

07/15 00:41

AI was supposed to replace human labor. It did the opposite. For the first time in history, humans are cheaper than software. And AI is creating more jobs than it eliminates. Technology has

中文介绍 博主gsivulka以「你刚雇佣了一百万个糟糕的员工」为引子，暗示了对AI工具或其应用效果的批判性观点。

How to Build a Self-Improving Outbound System on Codex

@nifinet · 11.0K 粉丝 · 337.0K 阅 · 518 赞 · 33 转

07/19 22:36

Earlier this year, Andrej Karpathy (@karpathy) pointed an agent at his own training code and let it run for two days. It ran 700 experiments, kept the 20 that beat the benchmark, and made the model

中文介绍 介绍如何构建一个自我改进的系统，灵感源于 Andrej Karpathy 的实验。Karpathy 让一个 agent 针对自身训练代码运行两天，执行了 700 次实验，并筛选出 20 个优于基准的成果。此方法展示了通过 AI agent 驱动的自动化实验，持续优化模型和代码的潜力。

GLM 5.2 With Vision

@part_harry_ · 1.5K 粉丝 · 225.6K 阅 · 503 赞 · 46 转

07/16 12:24

GLM 5.2 is one of the best currently available open source language models. However, unlike other flagship models like Qwen, Kimi and Minimax, GLM 5.2 does not support image inputs. We took this as a

中文介绍 博主讨论开源语言模型GLM 5.2的视觉能力（Vision），可能是在为其探索或开发图像输入功能。这一举措旨在增强GLM 5.2的多模态处理能力，使其能够处理视觉信息，从而拓宽应用范围并提升其在与Qwen、Kimi等模型竞争时的竞争力。

The Self-Driving Company

@amasad · 472.6K 粉丝 · 223.7K 阅 · 560 赞 · 65 转

07/17 01:07

We are beginning to see what happens when a company learns to operate itself. In the past six months, engineers at Replit have nearly tripled code output. Review times held steady. Reversions and

中文介绍 博主提出了「自驱动公司」这一概念，暗示企业运营未来将高度自动化，能像自动驾驶汽车一样自我管理和优化。这可能涉及利用AI和自动化技术，大幅提升公司的运营效率、决策速度和资源配置能力，预示着企业组织形态和生产力的一次重大变革。

Maybe Intelligence Ain't All That

@CliffordSosin · 16.4K 粉丝 · 197.0K 阅 · 514 赞 · 49 转

07/19 05:36

Superintelligence arrived. You probably didn't notice, because it turned out to be kind of incremental. Don't believe me? Run the test. Talk to Fable 5 for an hour, then talk to your ten smartest

中文介绍 博主探讨了「超级智能」的到来，指出其以渐进式而非颠覆性方式出现，因此可能未被察觉。他建议进行一项测试：与 Fable 5 交流一小时，再与十位最聪明的人交流，以感受这种「增量式超级智能」。

How OpenAI’s Sol Finally Learned Design Taste

@DesignArena · 17.6K 粉丝 · 191.5K 阅 · 509 赞 · 46 转

07/16 00:36

We benchmarked GPT-5.6 Sol on Design Arena’s Web Design (Non-Agentic) Arena, and we were surprised to find that it ranks 1st overall. This is 18 places higher than its predecessor GPT-5.5, and is the

中文介绍 DesignArena 提出了「How OpenAI’s Sol Finally Learned Design Taste」的问题，可能旨在探讨 OpenAI 的 Sol 模型在学习和掌握设计美学方面的进展和方法。

How to build your first team of agents:

@EXM7777 · 126.7K 粉丝 · 173.6K 阅 · 521 赞 · 51 转

07/18 21:48

I'm going to show you how to build your first team of AI agents... and how to wire them into a self improving loop inside a shared workspace the team we're building has 5 agents, one for each function

中文介绍 该推文详细讲解如何构建首个 AI 智能体团队，并将其整合到自我改进循环的工作流中。教程涵盖在共享工作空间内连接智能体的步骤，特别提到构建一个由 5 个具备独立功能的智能体组成的团队。

Graph Engineering with Claude: 14-Step roadmap from 0 to graph architect (Full Course)

@0xCodez · 23.9K 粉丝 · 154.5K 阅 · 510 赞 · 73 转

07/20 19:23

Most people who try to build a multi-step agent end up with a straight line. Step one, step two, step three - each waiting politely for the last to finish before it starts. 9/10 notice that half those

中文介绍 博主分享利用 Claude 进行「图工程」的 14 步路线图，旨在解决构建多步骤 agent 时常见的线性执行问题。许多人在设计 agent 时常陷入简单串联模式，而此教程将展示如何从零开始，通过图工程方法构建更复杂、非线性的 AI agent 工作流。

The Next Horizon in Agents

@btaylor · 158.6K 粉丝 · 144.3K 阅 · 511 赞 · 40 转

07/16 22:48

Introducing Horizon Today, Sierra is announcing Horizon, a platform that enables agents to pursue long-horizon goals, like originating a loan or getting prior authorization for a healthcare procedure.

中文介绍 Sierra公司发布了名为Horizon的新平台，旨在赋能AI Agent处理复杂且「长周期」的实际业务目标。该平台支持Agent完成如发起贷款或获取医疗程序预授权等多步骤任务。这标志着AI Agent从执行短期指令向处理更深层次、更具商业价值的复杂流程迈进，为企业自动化提供了新的可能性。

Gemini API Managed Agents: free tier, triggers & more

@GoogleAIStudio · 186.6K 粉丝 · 111.0K 阅 · 512 赞 · 51 转

07/17 01:05

Today we’re announcing new capabilities for Managed Agents in Gemini API, including free tier availability, budget control guardrails, and scheduled triggers. Building on our previous release of

中文介绍 Google AI Studio宣布Gemini API的Managed Agents新增功能，包括提供免费层级、预算控制保护措施以及支持定时触发器。这些更新建立在之前的发布基础上，旨在降低开发者使用门槛，赋予他们更大的灵活性来管理和部署AI Agent。新功能将促进更广泛的Agent应用，并帮助用户更好地控制成本与调度。

A Framework for Frontier AI and the Dawning of a New Age

@demishassabis · 1.4M 粉丝 · 98.1K 阅 · 1.5K 赞 · 236 转

07/14 17:10

This is a pivotal moment in human history. Artificial General Intelligence (AGI), a system that exhibits all the cognitive capabilities the brain has, is probably only a few short years away. When we

中文介绍 Google AI CEO Demis Hassabis认为，人类正处于关键历史时刻，通用人工智能（AGI）的实现可能仅在数年之内。AGI系统有望展现人类大脑的全部认知能力。他可能在分享一个关于前沿AI的框架，以应对即将到来的新时代。

How we built our knowledge base

@cerebras · 60.4K 粉丝 · 70.7K 阅 · 509 赞 · 38 转

07/17 02:27

Authors: @hi_im_isaac_, @learnwdaniel, @gaozenghao note: the interactive version of full technical blog available: https://www.cerebras.ai/blog/how-we-built-our-knowledge-base Employees ask our internal knowledge base more than 15,000

中文介绍 Cerebras发布技术博客，详细阐述了他们如何构建企业内部知识库。该知识库每月处理员工超过15,000次的查询，极大提高了工作效率。博文由多位作者共同撰写，提供了实际案例和技术细节，为其他组织构建高效智能知识管理系统提供了宝贵的实践经验和设计思路。

Kimi K3 on KernelBench

@elliotarledge · 40.7K 粉丝 · 69.3K 阅 · 506 赞 · 45 转

07/17 17:24

This is the Kimi K3 post you guys have been waiting for. I got some early access to this model and have been testing it on kernels, and even before seeing the benchmark scores I was impressed with its

中文介绍 博主获得了Kimi K3模型的早期访问权限，并对其进行了初步测试和KernelBench基准评估。在未查看具体分数前，博主已对Kimi K3的性能印象深刻。这篇分享为期待Kimi K3的用户提供了第一手体验和性能洞察，预示着该模型可能在某些任务上展现出强大的能力。

I Brutally Modified My Front-End Design Skill ~ Now My UIs Don’t Look Like AI Crap

@JoePro · 949 粉丝 · 66.8K 阅 · 537 赞 · 29 转

07/14 11:51

To create a custom Claude skill, navigate to Customize > Skills in your Claude interface, click the + button, and select + Create skill

中文介绍 详细介绍了在 Claude 界面中创建自定义 skill 的具体步骤。用户需进入「Customize > Skills」，点击「+」按钮，然后选择「+ Create skill」即可。这为希望扩展 Claude 功能的用户提供了清晰的实践指导。

Kimi K3 Reverse Engineering - Opus Without The Nagging

@herrcore · 15.4K 粉丝 · 64.2K 阅 · 514 赞 · 84 转

07/20 03:28

tl;dr K3 makes us question why we are paying Anthropic. We’ve been using Opus as our go-to model for reverse engineering tasks since the release of 4.5, and with each new release, it feels like we

中文介绍 博主对 Kimi K3 进行了逆向工程任务评测，并与 Anthropic 的 Opus 模型对比。评测结果表明，Kimi K3 在此类任务中表现出色，甚至让博主质疑继续为 Opus 付费的必要性。这暗示 Kimi K3 可能是逆向工程领域中，一个性能与性价比俱佳的替代方案。

Let's build Claude Code's harness (step-by-step)

@akshay_pachaar · 281.7K 粉丝 · 53.7K 阅 · 516 赞 · 74 转

07/16 02:10

We will cover everything that goes into building a coding harness, the agent loop, planning, subagents, sandboxing, memory, and checkpointing, built step by step. If you've ever tried building your

中文介绍 博主分享了为 Claude Code 构建编码辅助工具「coding harness」的逐步教程。内容深入讲解了智能体循环、规划、子智能体、沙盒环境、内存管理和检查点等关键技术组件的实现细节。

From Loop Engineering to Graph Engineering?

@IntuitMachine · 62.3K 粉丝 · 50.7K 阅 · 552 赞 · 47 转

07/18 18:00

What the shift in AI agent architecture is really about Peter Steinberger just posted nine words that gathered thousands of likes: https://x.com/steipete/status/2078277297791189132 "Are we still talking loops or did we

中文介绍 该推文探讨了 AI 智能体架构的关键转变，从传统的「循环工程」（Loop Engineering）转向「图工程」（Graph Engineering）。文章引用 Peter Steinberger 的热门观点，提出对当前智能体设计范式的深刻反思。

How I use Linear to manage agents (and why we should all be building software factories)

@enginoid · 1.1K 粉丝 · 49.9K 阅 · 522 赞 · 37 转

07/19 22:31

Linear is probably my best-value subscription at the moment. I pay the sole user fee of $12/mo and the company agents have eaten through ~400 issues in the past month. I use Linear to organize my

中文介绍 博主分享如何利用项目管理工具 Linear 管理 AI agent 的工作流。他每月支付 $12 订阅费，其公司 agent 在过去一个月内处理了约 400 个问题。该分享展示了将 Linear 与 AI agent 结合以大幅提升生产力的方法，并倡导构建「软件工厂」的理念。

How to build a self-improving code review agent

@zachlloydtweets · 12.1K 粉丝 · 42.7K 阅 · 527 赞 · 57 转

07/16 00:20

This is the third post in my series on how to build a cloud software factory: Build an issue triage agent Build a spec writing agent [This post] Build a self-improving code review agent I’ll describe

中文介绍 该推文是关于构建「云软件工厂」系列文章的第三部分，核心内容是如何搭建一个自我改进的代码审查 agent。博主将详细描述其构建过程。此前系列文章已涵盖问题分类 agent 和规范编写 agent，本次分享进一步完善了自动化软件开发的工作流。

Did you have a good weekend? Wonderful. Meet Axxon One

@therealDeFlock · 82.7K 粉丝 · 12.2K 阅 · 500 赞 · 120 转

07/20 08:08

"Axxon One supports automatic tracking of people as they move between cameras". "automatic tracking of people"... Ideal for logistics, warehousing, and Safe City projects, this AI detector

中文介绍 博主介绍了 AI 产品 Axxon One。该产品支持「跨摄像头对人员进行自动追踪」，并配备 AI 探测器。它尤其适用于物流、仓储以及智慧城市等项目，旨在为这些领域提供高效且自动化的监控与人员追踪解决方案，以提升运营效率和安全性。

FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation

👍 0

07/17 08:00

Video Diffusion Transformers process long spatio-temporal sequences, making self-attention the main bottleneck in high-resolution video generation. Training-free sparse attention reduces this cost, but adaptive Top-p routing creates uneven per-head workloads under multi-GPU sequence parallelism. The

中文介绍 FVAttn提出一种自适应稀疏注意力机制，旨在解决视频扩散Transformer在处理长时空序列时自注意力成为高分辨率视频生成瓶颈的问题。它通过运行时负载均衡优化了现有训练无关稀疏注意力方法，以应对多GPU设置下Top-p路由导致的工作负载不均。

When Does Muon Help Agentic Reinforcement Learning?

👍 12

07/17 08:00

Muon is competitive with AdamW in large-scale pre-training, but its value for reinforcement-learning (RL) post-training remains unclear. We study vanilla Muon in sparse-reward agentic RL through matched single-seed comparisons with AdamW on ALFWorld using Qwen2.5-0.5B-Instruct. Under Group-in-Group

中文介绍 该研究探讨了Muon优化器在稀疏奖励的智能体强化学习（RL）后训练中的价值。通过在ALFWorld环境中使用Qwen2.5-0.5B-Instruct模型，对比了原始Muon与AdamW的性能，以明确Muon何时能助力RL。Muon在大规模预训练中与AdamW表现相当。

An Exam for Active Observers

👍 0

07/17 08:00

Human vision is a closed loop: gaze is continuously redirected by intermediate hypotheses rather than a single snapshot. Decades of psychophysics and cognitive science have argued that this active observation is essential for a wide range of tasks. Whether today's multimodal large language models (M

中文介绍 本文探讨了人类视觉的「主动观察」特性，即凝视会根据中间假设持续调整。心理物理学和认知科学认为这种主动观察对许多任务至关重要。研究旨在通过一项“考试”来评估当前的多模态大模型（MLMs）是否也具备类似的主动观察能力。

CADAQUES: A Cost-Aware Dual Architecture for Query-Efficient Autonomous Discovery

👍 0

07/17 08:00

Autonomous discovery systems couple a resource that answers queries (a simulator, instrument, or analytic model) to an algorithm that selects what to query next. Most software frameworks for this loop inherit the control structure of numerical optimization: campaigns run for a fixed number of iterat

中文介绍 CADAQUES提出一种成本感知双架构，用于查询高效的自主发现系统。该系统将查询响应资源（如模拟器或分析模型）与选择后续查询的算法相结合，以优化现有软件框架中普遍存在的数值优化控制结构，从而提高发现效率。

The Honest Quorum Problem: Epistemic Byzantine Fault Tolerance for Agentic Infrastructure

👍 0

07/17 08:00

State machine replication (SMR) and Byzantine fault-tolerant (BFT) consensus guarantee agreement despite a bounded number of arbitrary, colluding faulty participants. However, these guarantees rely on participants outside this set correctly executing the protocol's transition semantics. Agentic vali

中文介绍 本文探讨了智能体基础设施中的“诚实法定人数问题”，即拜占庭容错（BFT）共识虽然能保证在有限恶意参与者下的协议一致性，但其前提是多数非恶意参与者能正确执行协议语义。研究旨在为智能体基础设施提供认知层面的拜占庭容错。

Audio-Visual Flamingo: Open Audio-Visual Intelligence for Long and Complex Videos

👍 6

07/17 08:00

We present Audio-Visual Flamingo (AV-Flamingo), a fully open state-of-the-art audio-visual large language model (AV-LLM) for joint understanding and reasoning over audio, images, and long-form videos. Unlike prior AV-LLMs that primarily focus on short clips, AV-Flamingo is designed for understanding

中文介绍 研究团队推出了Audio-Visual Flamingo (AV-Flamingo)，这是一个完全开源的先进音视频大语言模型（AV-LLM）。与之前主要关注短视频片段的AV-LLM不同，AV-Flamingo专为联合理解和推理音频、图像及长篇复杂视频而设计，旨在实现开放的音视频智能。

Understanding Reasoning from Pretraining to Post-Training

👍 17

07/17 08:00

Reinforcement learning (RL) has become central to improving large language models (LLMs) on complex reasoning tasks, yet RL post-training is largely studied in isolation from the pretraining that precedes it. As a result, two basic questions remain open: (1) how do pretraining choices (model size, d

中文介绍 强化学习（RL）对提升大语言模型（LLMs）在复杂推理任务上的性能至关重要，但RL后训练与预训练往往被孤立研究。本研究旨在理解从预训练到后训练过程中推理能力的变化，探讨预训练特性如何影响后续的RL微调效果，以弥补当前知识的空白。

Let the Body Follow: Coupled Egocentric Control for Whole-Body Robot Teleoperation

👍 0

07/17 08:00

Whole-body teleoperation requires users to coordinate perception, manipulation, posture, and mobility across multiple robot components. This coordination is difficult because users must simultaneously control the robot's head, arms, torso, and base while maintaining task awareness and avoiding kinem

中文介绍 机器人全身遥操作要求用户同时协调感知、操作、姿态和移动等多个机器人部件，这一过程因需同时控制头部、手臂、躯干和底座而极具挑战。该研究提出一种耦合以自我为中心的控制方法，旨在简化用户的复杂协调任务，提升全身机器人遥操作的效率和易用性。

JoyNexus: Service-Oriented Multi-Tenant Post-Training for VLA Models

👍 1

07/17 08:00

The post-training of Vision-Language-Action (VLA) models is essential due to the diversity of simulators, robot embodiments, and task objectives. Existing compute services, whether offered as direct accelerator rental or batch-workload submission, typically allocate an exclusive set of GPU and CPU r

中文介绍 视觉-语言-动作（VLA）模型的后训练因模拟器、机器人实体和任务目标的多样性而至关重要。JoyNexus提出一个面向服务的、多租户后训练平台，旨在解决现有计算服务通常为VLA模型分配独占资源的问题，从而更高效地支持VLA模型的训练和部署。

Loop the Loopies!

👍 53

07/17 08:00

We present Loopie, the most powerful looped Transformer to date. The Loopie series consists of two Mixture-of-Experts (MoE) models: a 20B-parameter model with 2B active parameters and a 6Bparameter model with 0.6B active parameters. Looped Transformers have long faced a challenge: given an N-fold in

中文介绍 研究团队推出了Loopie，迄今为止最强大的循环Transformer模型系列。Loopie系列包含两个专家混合（MoE）模型：一个200亿参数（20B）模型（激活参数20亿）和一个60亿参数（6B）模型（激活参数6亿）。该模型旨在解决循环Transformer长期面临的挑战。

SciForge: An AI-Native, Multimodal Workbench for Scientific Discovery

👍 0

07/17 08:00

Scientific work increasingly spans heterogeneous artifacts -- papers, code, datasets, scientific file formats, model outputs, figures, manuscripts, and team decisions -- yet general-purpose AI assistants rarely preserve these objects as a coherent, auditable research state. We present SciForge, a mu

中文介绍 科学研究日益涉及论文、代码、数据集等多源异构工件，但通用AI助手难以将其整合为可审计的科研状态。SciForge提出一个AI原生、多模态的科学发现工作台，旨在统一管理这些异构数据，为科学家提供一个连贯、可追溯的科研环境，提升科学发现效率。

Presentation, Not Mechanism: A Render Confound in Deprecation-Aware Memory Evaluation

👍 0

07/17 08:00

AI systems increasingly retrieve from records that revise themselves: issue threads, encyclopedic histories, policy logs, and long conversations. The challenge is not only finding relevant evidence, but deciding which claims remain in force, which were superseded, and when to abstain. Structured mem

中文介绍 随着AI系统越来越多地从自我修订的记录中检索信息，挑战不仅在于找到相关证据，还在于判断哪些主张仍然有效、哪些已被取代。本研究揭示了“渲染混淆”问题，强调了在评估AI系统对过时信息处理能力时，表象而非机制可能会导致错误的判断。

Data and Learning Where it Matters for Contact-Rich Manipulation

👍 0

07/17 08:00

Learned policies trained end-to-end on large datasets often remain brittle in high-precision tasks and struggle with generalization. We find that these limitations largely stem from a lack of structure and focus in data collection. Our key insight is to leverage dense data collection only for the cr

DSWorld: A Data Science World Model for Efficient Autonomous Agents

👍 6

07/17 08:00

Despite strong capabilities in data understanding and decision-making, autonomous data science agents still heavily rely on trial-and-error workflows that involve expensive computation. This bottleneck motivates models that can anticipate the effects of data science operations before real execution.

ContinuityBench: A Benchmark and Systems Study of Stateful Failover in Multi-Provider LLM Routing

👍 0

07/17 08:00

In production large language model (LLM) deployments, high API availability guarantees do not equate to conversational continuity. When a primary provider experiences an outage or strict rate-limiting, naive stateless failover mechanisms successfully maintain uptime but silently discard conversation

Orbis 2: A Hierarchical World Model for Driving

👍 0

07/17 08:00

Current world models operate at a single level of abstraction, with most prioritizing perceptual fidelity while lacking the spatial reasoning and semantic understanding required for real-world downstream tasks. We present a hierarchical driving world model that factorizes future prediction across tw

Exo2EgoPose: Leveraging Exocentric Demonstrations for Vision-Language guided Egocentric 3D Hand Pose Forecasting

👍 0

07/17 08:00

Perceiving multimodal cues and forecasting fine-grained actions from an egocentric (Ego) perspective is vital for applications like robot manipulation. However, previous studies either rely mainly on under-informed visual inputs to predict coarse human motions or follow the VRM/VLA paradigm, which s

EgoExoMoCap: Distributed Ego-Exo Human Motion Capture

👍 0

07/17 08:00

Human motion capture from head-mounted devices (HMDs) offers a scalable way to acquire real-world human motion and interaction data, which is crucial for applications in embodied AI and VR/AR. Existing approaches focus on either egocentric body tracking, estimating the motion of the subject wearing

Agentic Synthesis against Counterexample-Supplemented Sketches

👍 0

07/17 08:00

Coding agents can fix a failing example without preserving the domain rule that made it fail, so later generations can repeat the same plausible mistake. We present agentic synthesis against counterexample-supplemented sketches, a repository-native method for systems whose governing policy is discov

Making Agent-Mediated Contributions Governable: A Project-Level Governance Manifest for Open-Source AI Collaboration

👍 0

07/17 08:00

Generative AI and coding agents are intensifying a central governance tension in open-source software (OSS): they scale contribution generation faster than maintainers can assess risk, evidence, and accountability. Existing responses improve agent-readability and traceability, but project rules must

GeoChrono: Benchmarking and Rethinking Long-Term Temporal Understanding in Remote Sensing

👍 0

07/17 08:00

Remote sensing offers an unparalleled vantage point for observing the Earth's long-term surface evolution, yet it demands that a model not only perceive land cover at isolated moments, but also track changes, memorize evolution histories, and reason across time and space. However, existing studies l

SkillNav: Score-Level Skill Intervention for Zero-Shot Object Goal Navigation

👍 0

07/17 08:00

Vision-Language Model (VLM) agents have advanced zero-shot object-goal navigation, yet single-frame reasoning leaves them without the cross-step behavioral awareness an embodied navigator requires, producing recurring failures such as dead-end stalls, in-room loops, and circuitous approaches to dete

Verified LLM-Driven Synthesis for Concept Design

👍 0

07/17 08:00

Concept Design structures software systems around concepts: user-facing, self-contained units of functionality with a focused purpose. Concepts are composed into applications using synchronization rules called reactions, which specify how actions in one concept trigger actions in others. This paper

Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents

👍 0

07/17 08:00

Large language model (LLM) agents are increasingly used for complex information-extraction tasks, yet it remains unclear whether agentic components such as reflection and memory lead to observable and controllable improvements over fixed LLM workflows. We study this question through conference-paper

AC-VLA: Robust Out-of-Distribution Action Execution via Compositional Learning

👍 0

07/17 08:00

Vision-Language-Action (VLA) models excel at end-to-end robotic manipulation but struggle with out-of-distribution (OOD) generalization when familiar sub-tasks are recombined in unseen configurations. We identify two mutually reinforcing failure modes: trajectory overfitting, where models overfit to

S1-Omni: A Unified Multimodal Reasoning Model for Scientific Understanding, Prediction, and Generation

👍 8

07/17 08:00

We present S1-Omni, a unified multimodal reasoning model for scientific understanding, prediction, and generation. AI for Science (AI4S) has advanced significantly through domain-specific models, tool-augmented LLMs, and scientific language models. However, model capabilities remain highly fragmente

Understanding Agent-Reactive Bugs at the Model-Harness Boundary: An Empirical Study of LLM Agent Issue Reports

👍 0

07/17 08:00

LLM agents span command-line interfaces (e.g., Codex) and agent frameworks (e.g., LangChain), integrating backend LLMs with harness code that parses model outputs, controls agent loops, and manages context. Both the harness and LLM-generated responses jointly shape an agent's execution. This archite

ToolVerse: Unlocking Massive Environments and Long-Horizon Tasks for Agentic Reinforcement Learning

👍 0

07/17 08:00

While LLM agents demonstrate strong reasoning abilities in compact and well-defined scenarios, they struggle to maintain robustness and effectiveness when faced with large-scale, diverse, and dynamic real-world environments that demand seamless tool integration. To address this gap, we introduce Too

Do Agents Dream of False Memories? Black-box Visual Attacks on Long-term Memory in Multimodal AI Agents

👍 0

07/17 08:00

Multimodal AI agents increasingly rely on persistent long-term memory to ground generation in past visual and textual episodes. We show that unconditional trust in visual data creates a critical vulnerability. We propose Lucid, a black-box adversarial framework that compromises multimodal memory pip

Think at 5 Hz, Act at 20 Hz: Asynchronous Fast-Slow Vision-Language-Action Inference for Closed-Loop Driving

👍 1

07/17 08:00

Large language models bring instruction following and scene reasoning to end-to-end driving, but their inference latency collides with the control rate a vehicle requires. Existing closed-loop agents hide this gap by invoking the model on alternate simulation ticks and replaying the previous command

LayerProof Mylar

07/21 04:58

Create launch-ready motion videos for your products

中文介绍 LayerProof Mylar是一款视频生成工具，旨在帮助用户为产品快速创建可用于发布、具备动效的短视频。它简化了产品宣传内容的制作流程，特别适用于社交媒体和营销场景，使用户能高效产出视觉效果丰富的发布材料。

Loova Ads Studio

07/21 04:58

Create AI ads that convert, not just look good

中文介绍 Loova Ads Studio是一款基于AI的广告制作工具，致力于帮助用户生成能有效实现转化的广告内容，而非仅仅关注美观度，从而提升广告效果。

Skippr AI

07/21 04:56

The live AI employee inside your product, serving every user

中文介绍 Skippr AI是一款嵌入产品内部的实时AI「员工」，能够为每位用户提供个性化服务和支持，旨在提升用户体验。

Replay QA

07/21 04:55

Replay QA tells you what is broken before your users do

中文介绍 Replay QA是一款先进的质量保证（QA）工具，其核心功能是在最终用户发现软件缺陷之前，主动检测并告知开发者存在的问题，从而提高产品稳定性。

Fuzzy AI

07/21 04:52

We warm your prospects before reaching out

中文介绍 Fuzzy AI旨在在销售人员接触潜在客户之前，通过AI技术进行「预热」，提升沟通效率和成功率。

Nautis

07/21 04:52

The AI-native Operating System for founders.

中文介绍 Nautis是一款面向创业者的AI原生操作系统，它集成了多种人工智能工具和功能，旨在为创始人提供一个集成的、智能化的平台，助力其业务发展与管理。

Scriptyard

07/21 04:50

A free visual story development workspace

中文介绍 Scriptyard是一款免费的视觉化故事开发平台，为用户提供一个直观的工作空间，以便于规划、创作和组织故事内容。

CaptureKit

07/21 04:42

Keep your desktop clean by auto-organizing screenshots.

中文介绍 CaptureKit是一款实用的桌面工具，其核心功能是自动整理用户截取的屏幕截图，从而有效保持桌面整洁，并提升文件管理的效率。

Backbeat Forge

07/21 04:42

Turn drum audio into a score you can actually edit

中文介绍 Backbeat Forge是一款创新软件，能将输入的鼓点音频文件转换为可编辑的乐谱。用户可以通过此工具对生成的鼓谱进行调整和修改，从而简化音乐制作、编曲和鼓点编辑的工作流程，提升创作效率。

Creed

07/21 04:42

Your personal context file for every agent.

中文介绍 Creed为每个AI代理提供其专属的「个人上下文文件」，旨在帮助AI代理更好地理解和处理特定任务。

Devin AI: The Full Beginner’s Guide (Better Than Claude Code?)

07/17 02:25

中文介绍 这则YouTube视频提供了Devin AI的全面初学者指南，深入探讨了这款人工智能工具在代码开发方面的应用潜力。视频可能通过演示和比较，向观众展示Devin AI的功能和使用方法，并讨论其与Claude Code相比的优劣，以帮助初学者快速掌握。

Director by OpenArt

07/17 02:12

中文介绍 这则YouTube短视频介绍了OpenArt公司推出的一款名为「Director」的新产品或功能。该视频可能简要展示了「Director」的主要特点和用途，暗示其可能是一款用于创意或艺术领域的人工智能工具。

Codex and Claude Shipped Browser Updates. This Changes Everything.

07/15 21:00

中文介绍 Codex和Claude两大AI模型近期发布了浏览器更新，预计将对用户与AI的互动方式产生显著影响。这些更新可能包括功能改进、性能优化或更深度的浏览器集成，旨在提升用户体验并拓宽AI工具的应用场景。

OpenAI Just Merged ChatGPT and Codex. This Changes Everything.

07/13 01:24

Grok 4.5 + Cursor: Full Guide

07/10 01:51

Build Amazing Websites with Claude Fable 5 (Full Guide)

07/07 01:37

How to Create Your Own AI Assistant (It's So Easy)

07/01 09:40

Poweful Agents inside Slack

06/25 06:39

AI Agents Just Changed Forever: GLM 5.2, Codex Skills, Claude & Cursor

06/21 21:30

9 AI Agent Skills To Get Ahead of 99% of People

06/18 10:27

求购或者有偿用一次 dayone plus 账号

07/20 21:54

6 回复 · Apple 节点

某些 App 能不能不要这么自作聪明

07/20 21:47

11 回复 · 程序员 节点

有什么比较丝滑的远程桌面工具

07/20 21:32

16 回复 · 程序员 节点

AI 时代低代码产品是否还有意义？

07/20 18:14

19 回复 · 程序员 节点

土区 App Store 内购方式的 Youtube Premium 被暂停， renew 要求绑定支付方式

07/20 14:17

15 回复 · Apple 节点

土区内购家庭组 YouTube，现要求添加付款方式或手机号，大佬指点下

07/20 11:08

7 回复 · Apple 节点

iPhone 无法批量导出视频是什么鬼？

07/20 06:59

6 回复 · Apple 节点

iOS 如何设置能实现像安卓手机的来电提醒一样显示是谁打电话过来的，比如 10086 标识为中国移动这样？

07/19 11:28

14 回复 · Apple 节点

iPhone 换电池，都是去官方还是自己买电池换呢？

07/18 20:19

20 回复 · Apple 节点

强烈推荐一款性价比极高的 Linux 本搞 vibe coding

07/17 12:24

33 回复 · Linux 节点

Human mathematicians are being outcounterexampled

07/21 03:03

47 points · 7 comments

Safety and alignment in an era of long-horizon models

07/21 02:43

14 points · 2 comments

Making ASCII Art in Vim

07/21 02:42

10 points · 0 comments

Nativ: Run frontier open models locally on your Mac

07/21 02:16

80 points · 31 comments

The drivers behind software delivery inefficiency

07/21 02:10

27 points · 12 comments

Agent swarms and the new model economics

07/21 02:06

47 points · 16 comments

That post never existed. Stop listening to that thing

07/21 01:57

97 points · 88 comments

Hyprland 0.55 announced the switch to Lua for its config files

07/21 01:31

90 points · 71 comments

Kimi Work

07/21 01:13

198 points · 102 comments

Jelly UI: Soft-body physics for native HTML form controls

07/21 01:07

185 points · 73 comments

How we measured AI writing across arXiv, and where the measurement breaks

07/21 00:36

169 points · 122 comments

Launch HN: Bloomy (YC S26) – AI-powered mastery learning for K-12

07/21 00:32

Hi HN, I’m Alex Southmayd, the founder of Bloomy (https://bloomylearning.com) – an AI-powered mastery-learning platform for K-12 students. Bloomy provides students with an AI tutor alongside adaptive curriculum (right now Math, English Language Arts, and Writing).How it works: we diagnose

I Stopped "Creating Content"

07/20 23:47

126 points · 89 comments

Ziggity – A terminal UI for Git, written in Zig

07/20 23:46

54 points · 34 comments

Cross sectioning insects in an electron microscope with a femtosecond laser [video]

07/20 23:40

47 points · 2 comments

The Voice of Google

07/20 23:15

117 points · 46 comments

Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

07/20 23:13

223 points · 233 comments

Corners Don't Look Like That: Regarding Screenspace Ambient Occlusion (2012)

07/20 23:07

129 points · 51 comments

Controlling Reasoning Effort in LLMs

07/20 22:35

53 points · 2 comments

China's open-weights AI strategy is winning

07/20 22:21

699 points · 600 comments

Perfection Is Not Over-Engineering

07/20 22:10

154 points · 74 comments

Firefox Merges Support for Vulkan Video Decoding

07/20 21:47

205 points · 50 comments

Shinjuku Station in 3D

07/20 21:43

104 points · 19 comments

Hacker wipes Romania's land registry database

07/20 21:28

496 points · 268 comments

LEDs' Potential to Save Our Night Skies

07/20 21:07

177 points · 130 comments

The EU is about to sell our most sensitive data to the US for visa-free travel

07/20 20:14

443 points · 248 comments

Airport Simulator

07/20 18:30

572 points · 118 comments

Exploit brokers pay $500k for WordPress RCEs. I found one with GPT5.6 and $25

07/20 16:13

358 points · 203 comments

ECC and DDR5

07/20 00:31

100 points · 95 comments

85.3 GFlops: Optimizing FP32 Matrix Multiplication on a Single AMD Zen 3 Core

07/17 22:24

19 points · 8 comments

v2.1.215

07/19 10:56

What's changed Claude no longer runs the /verify and /code-review skills on its own; invoke them with /verify or /code-review when you want them

中文介绍 Anthropic公司旗下Claude代码工具发布v2.1.215版本更新。此次更新调整了"/verify"和"/code-review"技能的运行方式，它们不再自动执行，用户需要时需通过指令明确调用。

v2.1.214

07/18 09:20

What's changed Fixed single-segment dir/** allow rules like Edit(src/**) auto-approving writes to nested dir/ directories anywhere in the tree instead of only /dir Fixed a permission-check bypass affecting commands run in Windows PowerShell 5.1 sessions Fixed Bash permission checks to fail closed on

中文介绍 Anthropic Claude Code发布了v2.1.214版本。此更新主要修复了多项错误，包括单段目录通配符允许规则、影响Windows PowerShell 5.1会话的权限检查绕过问题，以及Bash权限检查错误。

v2.1.212

07/17 08:26

What's changed /fork now copies your conversation into a new background session (its own row in claude agents) while you keep working; the in-session subagent it used to launch is now /subtask Added claude auto-mode reset to restore the default auto-mode configuration, with a confirmation prompt (pa

中文介绍 Anthropic 的 Claude Code 项目发布了 v2.1.212 版本。主要更新包括："/fork" 命令现在会将当前对话复制到一个新的后台会话中，用户可继续工作；原有的会话内子代理功能则被新的 "/subtask" 命令取代。此外，还新增了 "claude auto-mode reset" 功能，用于恢复默认的自动模式配置。

v2.1.211

07/16 07:02

What's changed Added --forward-subagent-text flag and CLAUDE_CODE_FORWARD_SUBAGENT_TEXT environment variable to include subagent text and thinking in stream-json output Fixed permission previews relayed to chat channels not neutralizing bidirectional-override, zero-width, and look-alike quote charac

v2.1.210

07/15 07:45

What's changed Added a live elapsed-time counter to the collapsed tool summary line so long-running tool calls visibly tick instead of looking stuck Added a startup warning for Write(path), NotebookEdit(path), and Glob(path) permission rules — use Edit(path) or Read(path) instead Fixed isolation: 'w

v2.1.209

07/14 14:36

What's changed Fixed /model and other dialogs being blocked in claude agents background sessions (reverts an overly broad guard)

v2.1.208

07/14 09:10

What's changed Added screen reader mode: opt-in plain-text rendering for screen reader users. Run claude --ax-screen-reader, set CLAUDE_AX_SCREEN_READER=1, or add "axScreenReader": true to settings. Added vimInsertModeRemaps setting: map two-key insert-mode sequences like jj to Escape in vim mode Ad

v2.1.207

07/11 08:52

What's changed Auto mode is now available without CLAUDE_CODE_ENABLE_AUTO_MODE opt-in on Bedrock, Vertex AI, and Foundry; disable via disableAutoMode in settings Fixed the terminal freezing and keystrokes lagging while streaming responses containing very long lists, tables, paragraphs, or code block

v2.1.206

07/10 09:45

What's changed Added directory path suggestions to /cd, matching /add-dir behavior Added a /doctor check that proposes trimming checked-in CLAUDE.md files by cutting content Claude could derive from the codebase /commit-push-pr now auto-allows git push to the repo's configured push remote (remote.pu

v2.1.205

07/09 05:22

What's changed Added an auto mode rule that blocks tampering with session transcript files Fixed --json-schema silently producing unstructured output when the schema was invalid, and schemas using the format keyword being rejected Fixed a message sent while Claude was working being silently lost whe