参考文献¶
本书基于 1927 篇一手 AI 材料系统编撰。 涵盖 410 个来源站点。
主要来源¶
| 来源 | 文章数 |
|---|---|
| mp.weixin.qq.com | 848 |
| aws.amazon.com | 239 |
| theregister.com | 22 |
| interconnects.ai | 22 |
| oneusefulthing.org | 21 |
| huggingface.co | 21 |
| mp.weixi | 19 |
| thehackernews.com | 16 |
| netflixtechblog.com | 16 |
| arxiv.org | 14 |
| developer.nvidia.com | 14 |
| deepmind.google | 11 |
| anthropic.com | 10 |
| testingcatalog.com | 10 |
| cio.com | 9 |
| hackread.com | 8 |
| juejin.cn | 8 |
| blog.crewai.com | 7 |
| vercel.com | 7 |
| claude.com | 7 |
完整索引¶
mp.weixin.qq.com(848 篇)¶
- 100 年压缩到 100 天!红杉资本:这就是 AGI (2026-05-12)
- 从0开发大模型的17种Agent架构演进详细拆解
- 18岁高中生用ai挖出150万未知天体首批chatgpt原住民毕业
- 这个五一节我做了个总结,在 2 年多的创业里,我一共做了 25 个 AI 项目
- 2 小时,0 行手写代码,我用 Claude 做了一个生产级 VSCode 插件 (2026-05-10)
- 2026-llm-rl-algorithms-deeplog-imba-ppo-dpo-grpo-marl
- 2026年最值得关注的15款开发者工具,你用过几个? (2026-05-16)
- 3个人带100个AI程序员,一个月烧掉130万美元!OpenAI:钱我出
- 300万人在存的Claude提示词 (2026-05-16)
- ICLR 2026 I 英伟达 & 普渡大学用agent闭环实现文生3D
- 开源 AI 知识管理搭档 Obsidian + Claude Code 完整集成指南
- 别再死磕 Prompt 了,高手都在设计循环:7 种 Agent 架构 + Loop Engineering 中文主流视角
- 800行代码实现 Open Claw 的 Tool、消息总线、子Agent管理架构 (2026-05-10)
- CVPR 2026 Highlight|让AI像电影人一样「看」视频,8B小模型反超GPT-5与Gemini 3.1 Pro
- 腾讯研究院AI速递 20260430
- 为OpenClaw配置网盘空间的最佳实践
- Agent演化系列:五、从全局看Agent演化——为什么三条路线最终会汇合
- Agent架构关键变化:Harness正在成为新后端 (2026-05-10)
- 你的竞争对手可能没有员工,但它快要能自己融资了
- 智能体编排层中的上下文管理架构 (2026-05-01)
- agent-engineering-principles-architecture-practice (2026-04-28)
- Agent 时代,我们架构师应该学什么? (2026-05-09)
- Agent框架OWL原理详解 (2026-05-09)
- agent-guide-core-concepts-overview
- 一篇看懂 Agent Harness 的结构! — 12组件+7决策完整框架 (2026-04-20)
- Agent Harness 解析:智能体架构深度拆解 (2026-05-11)
- agent-harness-comparison-a3-subagent-2026-06-10 (2026-06-10)
- Agent Harness 可观测性:生产级 AI 项目必须补上的一课
- Agent Hooks:把 Agent 工作流变成可编程的
- 被裁了想转 AI Agent?先看面试官到底在筛你哪 7 样东西 (2026-05-07)
- 读完 Agent Loop 工程手册,我有 8 个还没想明白的问题
- Agent 记忆怎么设计才靠谱:Memory in the LLM Era 论文 4 模块框架 + 10 方案对比 + 新方法 F1 38.79 + 4 条工程原则
- Agent Memory 架构本质 (2026-04-27)
- agent-memory-architecture-past-influence-future-ruofei (2026-05-12)
- Agent 记忆注入实战:5 维框架 + 4 前沿论文(MemGuide/STITCH/ACE/Lost in the Middle)+ Context Burst 动态控制
- Agent 记忆存储方案深度洞察
- 晓斌:从 People-Oriented 到 Agent-Oriented Infra —— 意图驱动 + 代码沉淀的进化体 (2026-06-04)
- Agent核心技术概念与范式发生了哪些演变以及背后的思考
- agent principle architecture engineering practice
- Agent 可靠性的工程解法:从 Skillify 看持续改进机制 (2026-04-26)
- agent-runtime-7-responsibilities-secondcurve-2026
- Agent 自我改进的六条路 (2026-04-28)
- Agent skill 迭代式编写实战 — 淘天集团物流技术团队(其林)
- 从 0 到 1 教你写 Agent Skill,让 AI 懂你的"潜规则" (2026-04-24)
- agent-skills-comprehensive-survey
- Agent Skills会不会淘汰Coze、Dify、N8N等低代码平台?
- 《从零实现 Agent 系统》连载 01|Agent 系统是什么:问题空间与架构切片
- 微信文章:深度解析 Hermes Agent 如何实现"自进化"及其 Prompt / Context / Harness 的设计实践 (2026-04-24)
- agent-从能用到管好中间差了什么
- agent-开发范式演进从环境工程出发
- Agent 开发范式演进:从环境工程出发,“简化”多源实时上下文 (2026-05-12)
- agent-时代我们架构师应该学什么
- Agent 时代的生产力悖论:当协作本身成为最大的瓶颈 (2026-05-09)
- 开源 agentcanvas:读 Logfire 日志,一键可视化整个智能体工作流
- Harness工程火遍硅谷,AgentCore今天交卷! (2026-04-24)
- MCP · Skill · Agent · LLM · Harness — 一张图讲清:Agentic AI 系统如何真正落地 (2026-04-29)
- 《从零实现 Agent 系统》连载 01|Agent 系统是什么:问题空间与架构切片
- AgentMemory 解析:给 Coding Agent 装上本地长期记忆 (2026-05-19)
- AgentRun CLI v0.1.0 正式开源:一行命令运行您的托管 Agent (2026-05-09)
- 从单兵作战到团队协作:AgentRun 的多 Agent 生产级协作方案
- AgentScope Builder 快速体验:用 Harness 框架快速构建企业自进化智能体
- agentscope-java-2.0-enterprise-distributed-harness
- 首个 Java Harness Framework:AgentScope 把 OpenClaw 带到企业分布式场景
- Agent架构关键变化:Harness正在成为新后端 (2026-05-10)
- Agent框架OWL原理详解 (2026-05-10)
- AGI 之路,可能从一开始就走错了 (2026-05-08)
- AGI 之路,可能从一开始就走错了
- Agivar 录屏教学:清华非十科技桌面 Agent — 大脑小脑双层架构 + Jittor 推理引擎 + 2.3× 速度 + 三层确定性
- 腾讯研究院AI速递 20260429
- 腾讯研究院AI速递 20260506
- AI Agent 工程师学习路线:面向资深后端/大数据工程师的能力地图
- 十年老技术开发的 AI Agent 探索之路 (2026-05-10)
- 十年老技术开发的 AI Agent 探索之路 (2026-05-10)
- AI Agent架构设计工具数量的陷阱5个边界清楚的工具胜过20个模糊工具
- Google Agentic RAG 跨语料库框架:充分上下文智能体 + 5 阶段管线(FramesQA 90.1%)
- AI 内容创作,开始进入“画布 Agent”时代
- 芯片里最贵的不是计算,是搬运数据:前谷歌芯片架构师手把手教你读懂AI芯片
- 构建 AI 编程智能体的质量防线:5 个实用的代码质量控制机制
- ai coding guide tmall deep dive
- AI Agent 评测实战指南:5 维指标体系 + L1/L2/L3 准出分级 + 评测集三层设计 + LLM-as-Judge 三模式 + 6 大 Benchmark + Python 评测代码
- AI Coding 入门指南 - 如何更好地让AI真正帮你干活 (2026-05-16)
- AI的Context Layer
- 我们刚过了人类最后一个劳动节?AI新职业的八个变化
- AI 时代的 Git 版本管理最佳实践 (2026-04-28)
- 世界读书日|AI时代,我们该读什么?
- 面向 LLM 的架构设计:什么是真正的 AI Friendly 架构?
- 后端架构 AI Friendly 的标准与路径:面向无人值守开发时代的系统重构 (2026-06-15)
- AI 硬件迎来寒武纪时刻:百度智能云如何成为爆发催化剂
- ai-infra-auto-driven-skills-v0-bbuf-giantpanda
- AI Infra入门干货总结:大模型是如何高效推理的
- ai-infra-math-foundations-rmsnorm-softmax-causal-mask-sampling-binnnliu-2026-06-17 (2026-06-17)
- AI Native 公司 Dan Shipper / Every 30 人做出 100 人产出:层次化思维 + SaaS 六维能力框架
- AI 原生搭建指南:都在用AI,公司为什么更慢了?
- 让项目管理也AI Native —— 两个Git仓库干掉了周报、洞察和效能报表
- AI Native 时代 —— 研发组织何去何从
- 50 行 Codex 自我蒸馏 prompt 火了一周,真正值钱的是 5 行那条
- How to Build an AI-Native Startup
- AI 原生团队的脏乱差:CEO 数字分身失败案例与 AI 销售线索分配的兴衰 (2026-06-01)
- 什么是 AI Native,AI 原生团队如何搭建?
- 第一批「AI原生」本科生,要毕业了
- AI Native 时代 —— 研发组织何去何从
- 很多企业做完 AI PoC,为什么还是上不了生产
- ai-production-development-workflow-openspec-superpowers-gstack (2026-05-07)
- ai-skills-middleware-migration-android-harmonyos-taobao-2026
- 云端 Agent 基础设施两条硬经验:CreaoAI 联合创始人的状态/代码解耦 + 凭据隔离
- 【译】AI 时代真正的护城河,不是大模型 (2026-05-12)
- AI视频工具悄悄走到了第三阶段
- 业务 Agent 增强层架构:复用通用 Agent 基座,把业务能力做成可验证增强层
- ai-内容创作开始进入画布-agent时代
- AI 编程的下一场架构迁移:从代码检索,到上下文操作 (2026-05-16)
- ai4s-2026-h1-frontier-panorama-yinxi
- AIAIGC峰会嘉宾阵容 (2026-05-08)
- 让 Coding Agent 从黑盒到透明:阿里云 Agent 观测审计数据采集实践
- alibaba-aone-agentic-rd-mode-xiangbangyu (2026-05-08)
- 代码一旦生产出来,首先是负债 —— 一个CIO的AI效能实践复盘
- alibaba-devix-harness-ops-agent-7x24
- 阿里云 EventHouse 企业级 Agent 上下文构建五维框架
- 阿里云 AgentLoop:企业级智能体一站式自进化平台(4 环飞轮 + LoongSuite 84% 字段覆盖率 + Trace2Dataset 节省 90% + Agent-as-a-Judge 90% 一致率 + 记忆库/经验库)
- 用两行代码将 AgentRun 集成到你的应用
- 5 分钟上手 AgentRun:从注册到第一个 Agent 运行
- CIO 正在抛弃 AI 生码率:一场关于什么才算产研提效的实践复盘
- aliyun-cloud-native-api-gateway-gateway-api-guide
- 基础设施到 Agent 体验丨从 Claude Fable 5 看安全护栏的演进
- aliyun-cms2-cli-skill-5-scenarios-chenyanbing-2026
- 阿里云端到端业务需求专家 Agent:Multica 平台 + superai-* 技能集群 + TDD/pre-push 质量门禁
- aliyun-mse-ai-task-scheduling-agent-sandbox-cost-90-percent
- AlphaEvolve: A coding agent for scientific and algorithmic discovery (2026-05-01)
- AlphaEvolve交出一周年炸裂成绩单!AI自我改进不再科幻 (2026-05-10)
- amap-abot-earth-0.5-3d-native-world-model
- 智能时空思考Agent|告别'搜与推':高德如何用Proactive Agent重塑LBS主动服务?
- 最佳 Claude Code 配置:Andrej Karpathy 的 CLAUDE.md,134+k star了!
- ANOLISA v0.3:阿里 Agentic OS —— Agent 系统管家(4 层安全 + Token 节省 + 毫秒级快照) (2026-06-04)
- Nature | Anthropic:蒸馏过程潜意识传递行为偏好 (2026-05-01)
- Anthropic 官方生产级 Agent 最佳实践:12 个可复用的 MCP 设计模式 (2026-05-07)
- Anthropic 官方技能最佳实践:14 个可复用的 Agent Skills 设计模式 (2026-05-01)
- 数据级 Harness 视角:架构师 JiaGouX 解读 Anthropic 95% 数据分析 + 5 个反直觉边界
- anthropic-95pct-data-analysis-skill-stack-architecture
- anthropic-95pct-data-analysis-summary-189-chars
- Anthropic 官方技能最佳实践:14 个可复用的 Agent Skills 设计模式 (2026-05-09)
- Anthropic发布「AI原生创业公司」手册:涵盖全流程四大核心阶段,一人公司法典来了 (2026-05-18)
- OpenAI 的最强对手,离「AI Windows」又近了一步
- Anthropic最新博客:生物学Agent的瓶颈不在模型,而在数据基础设施
- Anthropic 最新播客:如何打造下一代 Claude
- anthropic-claude-cowork-task-boundary-5-signals-6-stages
- Anthropic 官方 Agent Harness 平台:Claude Managed Agents 完整指南 (2026-05-09)
- anthropic-claude-managed-agents-platform-launch (2026-05-08)
- Anthropic 研究:编程 Agent 用于社会科学研究
- Anthropic 发布 Computer Use 最佳实践
- Anthropic 上线「做梦」功能,让 Agent 越睡越聪明 (2026-05-07)
- anthropic-dynamic-workflows-ultracode-deep-research-lyuyuebannzi
- Claude背后全是大厂老兵!Anthropic工程团队1680人画像曝光
- Anthropic创始人行动手册:打造一家AI-Native创业公司(附36页中文PDF)
- Anthropic创始人行动手册:打造一家AI-Native创业公司(附36页中文PDF)
- 从 Anthropic 到 Google:Agent Skills 正在进入“设计模式”阶段 (2026-05-09)
- 1680 份员工履历揭开 Anthropic 招人底牌:基础设施老兵更吃香 (2026-06-XX)
- anthropic-institute-when-ai-builds-itself-jiagoux-interpretation
- anthropic-managed-agents-scaling (2026-05-07)
- Anthropic 最新博客:MCP 没死,它又来了 (2026-04-27)
- Anthropic 最新论文:阻止 AI 叛变的方法(Model Spec Midtraining) (2026-05-07)
- anthropic-mythos-glasswing-30days-vulnerability-report
- Claude思考黑箱终结了!Anthropic 祭出AI读心术:揭秘Claude的隐藏想法! (2026-05-08)
- anthropic-pm-jess-yan-managed-agents (2026-04-30)
- Anthropic 最新博客:Prompt Caching 是构建 Claude Code 的一切 (2026-05-06)
- Anthropic 官方 Agent Harness 平台:Claude Managed Agents 完整指南 (2026-05-10)
- anthropic-官方技能最佳实践14-个可复用的-agent-skills-设计模式
- anthropic-官方生产级-agent-最佳实践12-个可复用的-mcp-设计模式-v2
- Anthropic 官方生产级 Agent 最佳实践:12 个可复用的 MCP 设计模式 (2026-05-16)
- Anthropic 最新论文:阻止 AI 叛变的方法 (2026-05-12)
- Anthropic 联创:2028 年实现 AI 自我构建的概率超过 60% (2026-05-10)
- anthropic出手ai的内心独白曝光了
- Anthropic最危险路线图曝光: 无限记忆、多智能体! 硅谷AI终局仅剩双雄决顶 (2026-05-10)
- anthropic联创2028-ai自我构建概率超60
- ICML 2026 | APO:将多教师冲突转化为动态约束,破解多模态大模型推理对齐难题 (2026-05-13)
- APPO:阿里高德 AMAP-ML 把 Agent RL 信用分配细化到每个决策点,13 项基准 +4 分 (2026-06-16)
- ara-agent-native-research-artifact-37authors
- arxiv-2605-26099-ssm-attention-sleep-consolidation-cmu
- 为了让agent更安全的工作,有多少人操碎了心
- 你还在手动调代码,这个工程师的平台已经在自动进化了 (2026-05-13)
- 浏览器 Agent 的失忆问题:Autobrowse 如何让每次探索变成永久技能 (2026-05-08)
- 浏览器 Agent 的失忆问题:Autobrowse 如何让每次探索变成永久技能 (2026-05-08)
- AutoClaw 使用体验:自带 66 个 Skill、可接入聊天工具、安全性高 (2026-05-11)
- 这篇52页综述把AI做科研这件事,明明白白划成了L0到L4五个等级
- 从工具调优到决策智能:AutoResearch在营销增长中的AI Native实践
- 我把 Karpathy 的 AutoResearch 搬到了软件开发领域,效果炸了
- 百度智能云新一代AI机密计算实例:从CPU到GPU全链路可信 (2026-04-13)
- 网盘存量代码迁移实战:我们如何用三层架构管住 AI 的输出
- baidu-unlimited-ocr-rswa-attention
- ERNIE 3.0→5.0:百度文心大模型的后训练进化 (2026-05-01)
- Boris Cherny 新访谈:开发工具正在从 IDE 变成 Agent 控制台
- boris-cherny-新访谈开发工具正在从-ide-变成-agent-控制台-v2
- Boris Cherny 新访谈:开发工具正在从 IDE 变成 Agent 控制台 (2026-05-16)
- 基于浏览器请求录制与AI代码生成的E2E接口自动化测试实践 (2026-05-19)
- 来自字节跳动TRAE的Harness Engineering指南 (2026-05-07)
- Cat Wu: Anthropic Claude Code/Cowork 产品负责人访谈 (2026-04-27)
- CDP Bridge MCP:让 LLM 操作真实浏览器 (2026-05-12)
- ChatGPT官宣26位未来之星 (2026-05-08)
- ChatGPT 官宣 26 位未来之星,他们是:穿墙少年、街头摊贩、盲童的朋友…… (2026-05-12)
- 破案了!为啥ChatGPT老想着「稳稳地接住你」
- ChatGPT默认模型大升级,GPT-5.5 Instant正式上线:新增记忆来源功能 (2026-05-16)
- 所有实验室都怕字节,所有人都在夸DeepSeek!美国研究员36小时中国AI行
- 从 Chroma 换成 Qdrant,我踩了 100 万向量的坑
- 深入解析Chromium的 AI Coding 开发体系
- 腾讯混元 CL-Bench Life:让大模型读懂你的日常生活 (2026-05-01)
- claude-agent-sdk-skills-reusable-knowledge
- Claude写代码错误率从41%降到11%:Karpathy的4条规则为什么不够
- claude code 20000 char source analysis
- claude-code-7-layer-memory-architecture (2026-05-07)
- Claude Code 的 Agent 工程 (2026-04-25)
- Claude Code Agent Teams 实战:怎么拆任务、控权限、收证据
- 看完 Claude Code Agent Teams,我更确定接下来拼的是 Agent Runtime,技术拆解:Lead、Task List、Mailbox 和 Hooks 是什么东西
- claude-code-agent-view-huashu (2026-05-13)
- 深度拆解 Claude Code:12 个可复用的 Agentic Harness 设计模式 (2026-05-10)
- Claude Code 源码拆解:从启动到多 Agent 扩展层 (2026-04-24)
- claude-code-best-community-fork-evolution-vibecoder
- claude-code-commands-usage-guide (2026-04-30)
- Claude Code 上下文工程 —— Anthropic 团队的工程实践 (2026-06-04)
- 两万字详解Claude Code源码核心机制
- 我用 Claude Code 做需求调研,像多了一个产品经理 (2026-05-06)
- Claude Code 从 Demo 到产线 · 企业 Harness 工程化的 8 道关卡
- claude-code-dynamic-workflows-8th-translation-xingxiaozhao
- Claude Code Dynamic Workflows 第 7 译本(架构师 JiaGouX 视角:任务级 Harness 统一框架) (2026-06-04)
- Claude Code Dynamic Workflows:把复杂长时任务交给一群 agent
- Claude Code 动态工作流源码解析:pi-dynamic-workflows 架构实现
- Claude Code Dynamic Workflows 实战模式与构建技巧(Thariq Shihipar 一手经验) (2026-06-04)
- Claude Code Dynamic Workflows 第 6 译本(玉澄 / 51CTO 视角,2026-06-04) (2026-06-04)
- 撕开Claude Code真相:让它好用的98.4%,是工程不是AI (2026-05-01)
- 最新Claude Code创始人:编程已经解决了,Harness重要性持续降低,CC未来只有100行代码
- Claude Code Harness 深度分析 (2026-05-12)
- claude code harness deep understanding
- claude-code-kairos-paradigm-2026 (2026-05-07)
- claude-code-large-codebase-agent-harness-13-patterns-tuutuiagi
- 大厂代码库几百万行,Claude Code怎么跑起来的?Anthropic首次公开全套打法 (2026-05-19)
- 面向大型代码库的 Claude Code 团队落地经验与扩展策略(Agent Harness)
- Claude Code 实践:token 效率提高 71.5 倍的工作流 (2026-05-13)
- claude-code-multi-agent-harness-source-analysis
- 让你的 Claude Code 满血复活,Anthropic 在 GitHub 上开源了个插件
- claude code openclaw memory comparison
- claude-code-personal-learning-system-ruofei
- fb134668f09a3b45c1813781f912ae4e7e26294d3b60332606983b946944c328 (2026-04-30)
- 【图解】Claude Code 源码解析 |Prompt 提示词模块 (2026-05-08)
- claude-code-routines-proactive-agent
- 原始文章存档 (2026-04-30)
- 复制这套神仙配置,让Claude Code全自动修Bug!告别每天重复教AI写代码
- 使用Claude Code:session管理与1M上下文 (2026-05-09)
- Claude Code 七种自定义方法:CLAUDE.md / Rules / Skills / Subagents / Hooks / Output Styles / System Prompt
- claude code skills mcp rules source analysis
- claude-code-skills-superpowers-practice (2026-05-02)
- Skills:让 Claude 记住「怎么做」,告别重复教学
- Claude Code 源码拆解:从启动到多 Agent 扩展层 (2026-04-25)
- claude code source deep dive warrior
- CLAUDE.md (2026-05-07)
- Claude Code Subagents 深度指南:上下文卫生实战
- Claude Code 1.0.24 删 .gitignore 又想清空 Redis,Anthropic 罕见致歉
- Claude Code 工具设计复盘(官方) (2026-05-07)
- Claude Code 之父最新访谈:编程已经结束、harness 将消失,Claude Code 将只有 100 行代码,loop 才是未来 (2026-05-14)
- 我把 Claude Design 做成了 Skill,人人都能成为顶级网站设计师 (2026-04-27)
- Fable 5 的信号:Agent 开始拼 Runtime — 架构师若飞的 Runtime Contract 框架
- Claude Fable 5 提示词泄漏:安全工程露出了底牌
- Anthropic 官方 Agent Harness 平台:Claude Managed Agents 完整指南 (2026-05-01)
- Claude Managed Agents 脑手分离升级:sandbox、vault 和状态层怎么拼起来
- claude-opus-47-并不是一次全面升级甚至部分能力大幅衰退nn大家应该在合适的场景下选择使用nn昨晚-opus-47-上线全网又炸了nn我仔细看了下官
- Claude统治一切!吞下这颗红药丸,焊工也是顶尖程序员
- Claude Skill 质检工具 Skill Craft (2026-04-30)
- 2 小时,0 行手写代码,我用 Claude 做了一个生产级 VSCode 插件 (2026-05-10)
- Claude 发布官方报告,承认存在 3 处质量退化问题 (2026-05-14)
- 打破SWE-bench唯分数论,首个独立测量harness的基准开源了
- CLI-Anything 中文实测:7阶段流水线 + 真实 Demo(Blender/FreeCAD/Draw.io)
- CLI系列④·选型CLI、MCP还是API? (2026-05-07)
- CLI、MCP 和 CLI+Skill,应该如何选?
- UIUC、Meta、斯坦福等最新综述!Code as Agent Harness
- 开源 Claude Code 本地代码知识图谱:code-review-graph 完整上手攻略 (2026-05-13)
- code-review-graph:Claude Code 本地知识图谱,减少 6.8 倍代码审查 Token (2026-04-25)
- Codeindex · 让大模型更好地理解你的代码 (2026-05-14)
- codex-autonomous-earning-money (2026-05-13)
- Codex /goal 实现拆解:长任务 Agent 不只是多跑几轮 (2026-05-14)
- codex-goal-source-code-deep-dive
- OpenAI Codex 三件套:6 个职位插件 + Sites 网站生成 + Annotations 精准批注 (2026-06-04)
- codex-sites-cloudflare-worker-one-click-deploy-geekhome
- Coding Agent 下半场:从个人提效到组织级研发体系 — AgentScope Harness + Open SWE + Minions/Inspect/Cloudbot 收敛
- Cognitive Alpha Mining via LLM-Driven Code-Based Evolution (2026-05-01)
- Cola DLM:字节跳动连续潜空间扩散语言模型
- 一个理想的组织,是AI Agent在组织人|AI跃迁者调研
- 你的AI代码越写越乱,他72小时合了14个PR每个都更好——差距只在一个机制
- 你的AI代码越写越乱,他72小时合了14个PR每个都更好——差距只在一个机制
- 从 30 分钟手搓 Agent,到 Harness 成为"新后端" (2026-05-09)
- 从 Anthropic 到 Google:Agent Skills 正在进入"设计模式"阶段 (2026-05-10)
- 上下文工程 - 三种Memory方案对比
- cordon-semantic-transactions-tool-using-llm-agents-2026-06-17 (2026-06-17)
- 扣子 3.0 离谱更新:把 Codex、Claude Code 拉进一个项目工作?
- 扣子 3.0 多 Agent 协同实战:指挥所有 Agent 的 Agent + 5 人团队 6 步流水线
- 扣子 3.0 正式发布:@ 一下全员开工 — 项目化 + Agent 编排 + 工具链打通的协作系统 (2026-06-04)
- 优秀的垂类Agent构建实战:CPU缓存与Agent上下文管理 (2026-06-12)
- cuhk-slim-skill-lifecycle-agentic-rl-arxiv-2605-10923
- cursor harness model production floor
- 从Cursor返聘归来,90后华裔女高管带Claude开启日更模式:token成本比工程师工资低多了! (2026-04-30)
- CVPR 2026 Highlight | 清华打破多模态音频生成的「通才困境」:Omni2Sound 音频基础模型开源! (2026-05-08)
- CVPR 2026 Highlight|让AI像电影人一样「看」视频,8B小模型反超GPT-5与Gemini-3.1-Pro
- CVPR冠军代码开源:小米SVOR破解视频消除三大顽疾,连人带影一键抹除
- 达尔文.skill 2.0正式开源发布!让你的所有skill左脚踩右脚实现自我进化
- 深势科技携手阿里云AgentRun (2026-05-07)
- DeepSeek MoE 并行策略与通信优化
- deepseek-v4-flash-pro-通往百万级上下文与万亿参数推理的新纪元-v2
- DeepSeek V4 (Flash & Pro) :通往百万级上下文与万亿参数推理的新纪元 (2026-05-16)
- DeepSeek V4是怎么训练出来的?58页论文深入解读 (2026-04-24)
- 【DeepSeek V4】Triton FP4 优化实战:近距离感受老黄的刀法 (2026-05-07)
- DeepSeek-V4深度拆解:一篇论文同时做了五件大事
- DeepSeek Thinking with Visual Primitives 深度解读
- deepseek视觉原语论文当所有人在堆图像分辨率时它在堆指代精度-v2
- DeepSeek视觉原语论文:当所有人在堆图像分辨率时,它在堆「指代精度」! (2026-05-16)
- DeerFlow · Hermes · OpenClaw 架构区别深度对比 (2026-05-06)
- Demis Hassabis YC 专访:Agent 才刚刚开始,AI 下一步是创造虚拟细胞 (2026-05-01)
- DESIGN.md:Google Stitch 与 VoltAgent 开源,AI 设计 Agent 的视觉约束对齐文件
- 多轮 Agent 场景下,滴滴的 EAGLE-3 训推加速实践
- didi-ibg-customer-experience-llm-quality-inspection-3-pipelines
- 基于钉钉机器人的 Qoder CLI / Claude Code 双引擎 AI 助手实践
- 豆包 Seed 2.0 Lite升级:给 Agent 装上眼睛和耳朵 (2026-05-06)
- ECC:一套给 AI Coding Agent 用的 Harness 操作系统
- 告别「工具人」时代:电商 AI 操作系统崛起 — All in One 跨平台 + 行业 KnowHow Skill 化 + 5 巨头布局
- EdgeClaw:端云两栖龙虾框架 (2026-04-25)
- 何恺明首个语言模型:105M参数,不走GPT自回归老路 (2026-05-13)
- 从提示词到控制框架,自主编码智能体的 7 个设计要素 (2026-06-14)
- Rod Johnson 回归:Embabel 与「最后一波由人类选择的框架」 — Spring 创始人对企业 AI Agent 框架的深度访谈 (2026-06-04)
- embodied-intelligence-sim-to-real-active-inference-behavior-tree-intrinsic-motivation-chenzhiyan-2026-06-17 (2026-06-17)
- 从语言涌现到协作涌现:如何让 AI 产生高质量决策
- 下一代企业数字化架构:系统CLI化、流程Skill化、员工Agent化
- Vibe Coding in Production — Erik Schluntz / Anthropic (2026-05-07)
- Video Agent 范式迁移与算力-人才飞轮:Ethan He 从 Cosmos 到 Grok Imagine 的第一手洞见
- Evals 到底在评什么?一文拆解 AI 评估的三种方法
- 一个 Mission 跑 16 天、烧 7.78 亿 Token:Factory 公开了多 Agent 系统的构建哲学
- 一个 Mission 跑 16 天、烧 7.78 亿 Token:Factory 公开了多 Agent 系统的构建哲学 (2026-05-14)
- fanling company as agent ai org reflection v2
- FastAPI APIRouter:如何正确组织路由 (2026-05-10)
- FastAPI上线实战:认证、限流、零停机,一套代码搞定 (2026-05-09)
- FastAPI上线实战:认证、限流、零停机,一套代码搞定 (2026-05-10)
- 飞书aily:企业级Agent智能伙伴 (2026-04-25)
- flashlabs-vertical-ai-startup-pivot
- 全球首个完全AI编写的训练框架:面壁ForgeTrain速度反超英伟达Megatron,年底要把国产算力软件重写一遍
- 软件,Agent 才是主用户:Foundation Capital Agent 时代六条洞察 (2026-05-08)
- Agent 如何管理其他 Agent:四种Sub Agent 模式
- 从 Prompt 到 Harness:Claude 官方学习资料 (2026-05-07)
- fudan-peking-ahe-agentic-harness-engineering (2026-05-07)
- 高德广告工程 AI Native 全流程提效实践:Spec/知识/验收/状态/Skills 五类工程产物 + K/S/T 知识底座 + Agent Team 平台化 (2026-06-15)
- AI伴行技术解析:基于空间智能的高可用Agent架构实践 (2026-05-14)
- 高德 AI-Native 生产线(第 3 期):7x24 Self-Healing Pipeline + Agent 自进化 (2026-06-04)
- gaode-marketing-autoresearch-ai-native-practice
- 高德路线规划演进:从工业引擎到端到端时空基座(MobilityBench + TransitLM 双路线)
- 告别"氛围编程":基于 Harness 治理和 SDD 的团队级 AI 研发范式演进与实践 (2026-05-07)
- 面向复杂算法任务的 AI Agent:Long-Running Harness 架构设计与模型迭代应用
- 你的AI代码越写越乱,他72小时合了14个PR——差距只在一个机制 (2026-05-13)
- AI Agent 架构设计(八):Gateway 架构设计(OpenClaw、Claude Code、Hermes Agent 对比) (2026-05-07)
- 捅破个人AI天花板!YC总裁开源GBrain:8层架构打造AI第二大脑
- YC 总裁开源了自己亲手写的 AI Agent 大脑,1 周就 1 万点赞。 (2026-04-27)
- GBrain v0.42.44.0 源码深度解析:5 大设计决策与工程可靠性(术哥无界视角)
- Google出手统一全模态检索:Gemini Embedding 2把文本、图片、音频和视频压进同一向量空间
- Gemma 4 12B:Google 多模态本地模型 —— 扔掉编码器 (2026-06-04)
- 摒弃冗长技能文档,Gene 用轻量化策略提升 Agent 效果
- 蛰伏一年,周衔团队带来首个具身基础模型,烹饪做实验弹琴,效果炸场 (2026-05-07)
- 用两个 Git 仓库 + AI 编码助手做项目管理(2026版)
- gitlab-14pct-layoff-agent-platform-ai-2026q1
- 打造 AI 智能体专属的代码知识库:GitNexus 完整上手攻略 (2026-05-09)
- glm5-scaling-pain-inference (2026-04-30)
- Google Open Knowledge Format (OKF) v0.1:让 AI 知识库有了通用格式
- Google 开发者福利,每月免费领取 10 美金,别忘了来领啊! (2026-05-16)
- gpt-54-烧完额度后我把七家国产-ai-公司-coding-plan-对比了一遍想不到最应该买的竟然是这家
- GPT-5.5全球首破!0源码盲写程序,编程AI进入新纪元 (2026-05-13)
- GPT-5.5实测有点翻车。\n\n写完文章后,我已经拿Codex中的GPT-5.5测了不少长步骤的复杂任务:做动画、做PPT、做网站。\n\n我的感受是这是个非常无聊的学霸,会做题会尽量达成你布置的任务,但做出来的东西就是相当无聊。在指令相同,调用Skill相同的情况下,还是Opus4.7出来的东西更符合我审美。\n\n我猜可能是OpenAI真的在很努力跑分了,所以模型也在做题这事上优化得过度。以及,在GPT-5.5的System Card报告里,GPT-5.5有29%的情况会撒谎说自己完成了不可能完成的编程任务,远高于GPT-5.4和5.3版本。 (2026-05-16)
- GPT-5.5来了!我撤回了退订ChatGPT的决定 (2026-05-11)
- GPT-5级推理能力塞进语音模型,OpenAI把同传翻译成本砍穿地板价
- GPT-Image-2 完全指南!附大量玩法案例,顺便开源我的生图 Skill ~ (2026-05-10)
- gpt-image-2完全指南
- GPT -Image 2神级提示词分享 (2026-05-16)
- 开源 AI 编程可查询的软件工程知识图谱:Graphify 完整上手攻略
- gsd-get-shit-done-context-management-tool (2026-05-07)
- YC掌门人60天写了60万行代码:gstack开源 (2026-04-25)
- Hacker News 热帖:AI 会写代码了,为啥还要用 Python? (2026-05-13)
- harness-design-long-running-apps (2026-03-28)
- harness-engineering-10-step-practical-guide-2026
- Harness 工程 14 步路线图:从单 Agent 到自改进系统
- Harness 到底是什么?看看 OpenClaw、Hermes、Claude Code 的演绎吧 (2026-05-07)
- Harness Engineering:耗时一周,我是如何将应用的AI Coding率提升至90%的 (2026-05-09)
- Harness Engineering:耗时一周,我是如何将应用的AI Coding率提升至90%的
- AI 不缺智商缺纪律:一场 Harness 工程化实践 (2026-XX-XX)
- harness-engineering-alibaba-java-case-study (2026-05-07)
- harness engineering comprehensive guide conardli
- 深入浅出 Harness Engineering 之核心模式与理念
- 再看 Harness Engineering:真正要设计的不是约束,而是可删的工作现场
- 关于 Harness Engineering 你需要知道的一切(2026 版)
- Harness Engineering 概念框架 (2026-04-27)
- Harness Engineering 的未来——什么会消失,什么不会
- harness engineering long term agent tasks
- harness-engineering-loop-vs-harness-machine-heart-2026-06-17 (2026-06-17)
- harness-engineering-systematic-explainer (2026-04-30)
- Harness Engineering 从理论到实战:行为正确性死结 + 上下文腐烂 + 可驾驭性 + Ashby 定律
- harness-engineeringai-能在真正出事会炸的后端系统里写代码吗-v2
- Harness Engineering:AI 能在真正"出事会炸"的后端系统里写代码吗? (2026-05-16)
- Harness Engineering实践,做了一个平台让AI一晚上自动评测和优化你的系统 (2026-05-14)
- harness-pilot-claude-code-plugin-yangtong-2026-06-17 (2026-06-17)
- Harness如何支撑Agent在生产环境稳定运行?
- 基于 Harness + SDD + 多仓管理模式的 AI 全栈开发实践|得物技术 (2026-05-06)
- Harness 之后:Agent 可靠性的关键,是状态边界和失败闭环
- hermes-9-module-architecture-winty (2026-05-12)
- hermes-agent-12-layer-full-configuration-guide
- Hermes Agent 内部的 8 个 Loop,以及它们为什么会形成复利? (2026-06-12)
- hermes-agent-closed-learning-loop (2026-05-07)
- 深度解析 Hermes Agent 如何实现"自进化"及其 Prompt / Context / Harness 的设计实践 (2026-04-30)
- Hermes Agent 保姆级教程:一句话组建你的 AI 打工团队
- Hermes Agent /goal 长任务运行时架构拆解:状态持久化、Judge 闭环与自主续航 (2026-05-19)
- 万字保姆级教程:Hermes+Kimi K2.6 打造7x24h Agent军团 (2026-04-25)
- 长期运行的 Agent 怎么管:Hermes 治理分层与 5 张卡 (2026-06-01)
- Agent Loop 源码导读:一次 Hermes 任务的完整生命周期
- 拆解 Hermes Agent 的记忆系统:一个生产级 AI 记忆是怎么设计的
- 深度拆解 Hermes Agent 记忆系统:它修正了 OpenClaw 的哪层误区? (2026-05-10)
- 拆解 Hermes Agent 的记忆系统:一个生产级 AI 记忆是怎么设计的
- Hermes Agent 记忆系统深度拆解 (2026-04-30)
- Paperclip · AI 公司操作系统 · 第 5 篇(完结) (2026-05-07)
- Hermes Agent Operator 上手:把一个 Agent 养成可运营系统
- 拆完Hermes源码发现Agent自我进化不需要训练模型
- hermes-agent-self-evolving-source-analysis (2026-05-08)
- hermes-agent-skill-crossover-optimization-skillevolver-darwin
- hermes-agent-skills-source-code-analysis-shuge
- Hermes Agent 爱马仕的三级 memory,到底在记什么?
- Hermes Agent 工具系统实战解析:40+ 工具为什么不用配置表
- Hermes Agent 为什么火了?和 OpenClaw 龙虾比一比
- 我给Hermes配了4个Agent (2026-05-06)
- 给 Hermes 装上显微镜:Agent 执行全知道 (2026-04-27)
- 从一次任务到一次进化:完整拆解 Skill 创建、复用、修补链路
- hermes-self-improving-overview-winty (2026-05-12)
- Hermes新顶流Agent Skills闭环系统深度解析
- 手把手:用 Hermes Skills + Karpathy 的 LLM Wiki 让 AI 越用越懂你
- HiClaw 发布 v1.1.0,提供 Kubernetes 集群部署实现,支持 Hermes Worker 运行时 (2026-04-29)
- HiClaw 发布 v1.1.0,提供 Kubernetes 集群部署实现,支持 Hermes Worker 运行时 (2026-05-16)
- ingress-nginx已退役higress如何平滑替代
- 如何把经验装到Skills
- 推荐系统进入大模型时刻:昇腾 NPU 如何支撑千亿级生成式推荐落地
- huggingface-ai-agent-glossary-model-scaffolding-harness-tool-skill-subagent
- ICLR 2026 I 英伟达 & 普渡大学用agent闭环实现文生3D
- ICLR 2026 | 英伟达×普渡大学:用Agent闭环实现文生3D
- ICML2026 观点论文 — Transformer 图灵完备性高度依赖上下文管理 (人大魏哲巍团队)
- iii-dev-worker-trigger-function (2026-04-30)
- imclaw通过微信飞书操控claude-code-coodex-gemini-clipi-agent蜂群
- IMClaw:通过微信/飞书操控ClaudeCode/Codex/GeminiCLI/Pi Agent蜂群 (2026-05-10)
- AI 写前端 ≠ 设计 —— Anomaly 创始人对 Vibe Coding 哲学批判 (2026-06-04)
- Impeccable:把 AI 前端设计变成可检查的工作流 — 33.4k Star 开源项目深度分析 (2026-06-04)
- Boris Cherny 新访谈:开发工具正在从 IDE 变成 Agent 控制台
- is-grep-all-you-need-pwc-retrieval-harness-coupling
- james-cowling-engineering-philosophy-ai-era
- openJiuwen 开源 Jiuwen Symphony — 海量技能的精准发现与稳定协同(技能编排与分发系统)
- 蜂群Agent来了!openJiuwen社区发布JiuwenSwarm,引领Coordination Engineering新范式
- openJiuwen开源SwarmFlow — 从'能协作'到'稳稳地干完',开创蜂群可控协同新范式
- 面试官皱眉:"你的 Agent 跑了10轮之后还靠谱吗"——上下文漂移与工具调用幻觉深度解析
- Karpathy 怎么看 AI Agent(七):当程序员贡献的 bits 越来越少,什么技能还值钱 (2026-05-13)
- 我把 Karpathy 的 AutoResearch 搬到了软件开发领域,效果炸了
- Karpathy × Boris 访谈:Software 3.0 时代编程完整地图 (2026-05-07)
- karpathy-llm-wiki-v2-2026 (2026-05-07)
- Karpathy LLM Wiki V2 深度分析:记忆生命周期 + 知识图谱 + 混合检索 + 落地路线图
- Karpathy加入Anthropic后首讲:Vibe Coding已死,Software3.0来了
- Karpathy 最新访谈:从 Vibe Coding 到 Agentic Engineering
- Karpathy 最新访谈:从 Vibe Coding 到 Agentic Engineering (2026-05-10)
- Karpathy 最新访谈:从 Vibe Coding 到 Agentic Engineering
- karpathy-vibe-coding-to-agentic-engineering (2026-05-07)
- 省下1.25倍算力!Kimi这篇论文,可能改写所有大模型的训练方式 (2026-04-28)
- Kimi K2.5 架构创新 — 1T MoE 一层路由 + 三能力跃迁
- Kimi K2.6背后的Agent Database:Agent-native时代的数据Infra竞争,跟过去30年有何不同
- Kimi K2.6背后的Agent Database:Agent-native 时代的数据Infra竞争,跟过去30年有何不同 (2026-05-13)
- kimi-work-beta-foundation-model-company-advantage
- Kimi Work 发布 + Codex 并入 ChatGPT:通用 Agent 战场从云端迁移到本地 (2026-06-04)
- 快手首个打工人Agent (2026-05-08)
- Kuse.ai Junior:全球第一个AI员工 (2026-04-25)
- The Anatomy of an Agent Harness 解读
- LangChain用Agent做销售获客,3个月转化率提升2.5倍,看完后我发现,国内 Agent 落地的方法都错了 (2026-05-11)
- langgraph-a2a-adversarial-agent-team
- LangGraph 底层原理:它是怎么把 LLM 变成一台状态机的 (2026-04-27)
- laser acl2026 latent superposition visual reasoning
- 智能时空思考Agent|首个真实出行隐式意图评测基准LBS-IntentBench正式开源 (2026-05-07)
- 量子位编辑作者招聘
- LiveKit Agents:给大模型接上麦克风,没你想的那么简单
- LLM agent脚手架如何具备自进化能力?——以hermes agent为例 (2026-05-09)
- 大语言模型为什么能像人一样说话和思考
- LLM Post-Training全景指南:从RLHF到GRPO再到AgenticRL (2026-04-25)
- llm-rl中的熵-part-1-熵的调控
- llm-rl中的熵-part-2-熵对训练的调控
- Anthropic宣告「递归自我提升」时代到来,LLM如何实现自我进化?全景综述带你一探究竟
- 别把 LLM 提取的主题当成真实变量 — 因果分析中 '生成变量' 的 4 类错误与 5 问诊断清单
- LLM Wiki 架构解析:Karpathy 的 Markdown 知识库模式
- llm-wiki-obsidian-wiki-gbrain-self-organization-self-evolution (2026-05-13)
- local-vs-cloud-agent-onsite-context-debate-xingxiaozhao
- 长周期 Agent 详解:从 Ralph Loop 到可接管 Harness
- 阿里巴巴 & 蚂蚁 LoongSuite GenAI 可观测语义规范:从统一数据语言到规模化落地 (2026-05-12)
- 从个人生产力到组织能力:LoongSuite-Pilot×SLS 的 AI Coding 度量实践
- Loop Engineering,从 Prompt 工程师到 Loop 架构师的 14 步路线图 (2026-06-16)
- Loop Engineering:5 类循环 + 6 个生产硬边界 + 从 Loop 到 Graph
- Loop Engineering:不再写提示词,而是设计替你写提示词的循环
- Loop Engineering:从 AutoResearch 到 Claude Code——循环设计的第一性原理
- Loop Engineering 实践指南:在 Code Buddy 中构建自主循环系统
- loop-engineering-ifanr-popular-science-critique-2026-06-16
- AI编程又变天了!Claude Code之父、龙虾创始人同时力捧新范式
- loop-engineering-karpathy-autoresearch-eval-ruler-allentang-2026-06-16
- loop-engineering-langchain-four-layer-loopcraft
- 给产品经理的loop engineering
- Loop Engineering 实用指南:先写刹车,再写循环 (2026-06-15)
- 当AI圈开始聊Loop:提示词工程已死,但杀死它的不是新技术
- Loop Engineering 综合实战(三层结构 + 五要素 + 解剖 6 组件 + 4 模式 + 成本公式 + 三款产品 Loop 能力对比 + 组织准备度总表 + Ralph Loop 极简主义 + 三大风险)
- Loop Engineering 中被忽略的另一半:验证器比生成器更重要
- LoopWM:循环 Transformer 世界模型,10 亿参数在 ScienceWorld 上超越 Claude Opus 4.6
- 从 Spec 到损失函数 — 真正会用 AI Agent 的人已经在设计循环(Elvis Sun LFD/损失函数开发)
- 低代码 Agent、框架 Agent、自研 Agent,分别适合谁 (2026-05-12)
- 独家对话罗福莉:AI范式已然巨变! (2026-05-02)
- MAC(multi-agent-coding):Skills + Hooks 两层 Harness —— 完全委托 0-20% 的解法 (2026-06-04)
- MapSatisfyBench:首个以满意度为核心目标的地图智能体评测基准
- Markdown 不会过时 (2026-05-13)
- Martin Fowler 的 AI 研发提醒:非确定性进了研发链路,Harness 才真正开始承重 (2026-05-09)
- Martin Fowler 的 AI 研发提醒:非确定性进了研发链路,Harness 才真正开始承重 (2026-05-10)
- matt-van-horn-22-claude-code-hacks-everyinc
- Matt Pocock Skills 完整工作流:从需求澄清到架构整理
- Matt Pocock Skills:AI编程技能集合(grill-me/grill-with-docs/caveman)
- Memento-Skills:让 Agent 通过技能外部记忆持续进化 (2026-05-01)
- memento skills let agents design agents
- Memory in the LLM Era: Modular Architectures and Strategies in a Unified Framework
- Memory 不是 RAG:Agent 记忆的系统性框架 (2026-05-07)
- 给 10 万 Star 的 Hermes 装个记忆外挂,AI 终于能越用越聪明了 (2026-04-24)
- Meta Skill:Skill 之上编排 Skill 的抽象层(OpenSquilla) (2026-06-04)
- 微软 Agent Framework 全栈指南:从 Hello Agent 到生产托管(Python)
- microsoft-agent-framework-structured-output
- microsoft-agent-framework-tools-overview-provider-matrix
- Microsoft Build 2026:微软 AI 独立日 —— 7 款 MAI 模型 + Scout 智能体 (2026-06-04)
- Microsoft Build 2026 全景:7 款 MAI 模型 + Scout 智能体 + OpenClaw 登 Windows + NVIDIA 合作 (2026-06-04)
- 小米 MiMo Code 开源 — 终端编程 Agent 三大主线(计算/记忆/进化)+ 与 Claude Code 工程分化
- Mind Lab LoRA 持续学习体系:δ-mem + MinT + LoRA Scaling Law + Macaron-A2UI
- 清华系团队出手!一张4090即可「爆改」,1.3B小钢炮震撼开源 (2026-05-13)
- 不用人类手写训练框架了!AI自己写代码,训出1B端侧「小钢炮」
- 一个 AI 还是不够的:MiniMax Agent Team(Mavis) (2026-05-13)
- MiniMax M2.7:开启模型的自我进化 (2026-04-25)
- 国内首个 Frontier 三件套开源大模型:MiniMax M3 完整技术拆解
- 大模型不认识马嘉祺?我们做了一次全链路排查
- AI4S 突破:深度原理 MIRA + MPA 材料基座模型 40 项实验全面 SOTA,递归自训练打造最强材料基座
- MiroFlow:Deep Research Agent 脚手架 —— 与 Code Agent 的 6 大工程差异 (2026-06-04)
- MNN-Sana-Edit-V2:端侧运行的图像漫画风编辑大模型 (2026-05-11)
- 中科院开源 MobileGym:浏览器内高并发安卓仿真平台,Mobile Agent 训练与评测基础设施
- model-harness-fit-agent-harness (2026-05-07)
- MRAgent:记忆是重建的,不是检索的——Cue-Tag-Content 图上多轮主动重建
- MSA上桌,稀疏注意力玩家开始三国杀:MSA/DSA/MoBA/NSA 四方案深度对比 (2026-06-12)
- Multica — 开源 Managed Agents 平台 (2026-05-07)
- 字节跳动 MUSE-Autoskill:自进化 Agent 五阶段技能生命周期,自生成技能超越人类上限
- 微软CEO长文提出新概念「token资本」:AI时代企业生死的关键
- 4000行代码撑起一个Agent框架?nanobot架构深度解析
- ICML 2026|告别「单线程」思维,智能体进化出了原生的并行推理大脑
- nature-ai-scientific-assistant-google-futurehouse
- NeurIPS 2026 使用闭源 AI 检测器 Pangram 批量 desk-reject 论文事件 (2026-06-04)
- 新程Alpha认知模型:4B参数端侧部署,群体智能以小搏大比肩GPT-5.4
- 你写的 Skill,及格了吗? (2026-05-09)
- 大模型社会智能如何诊断?浙大团队提出 NICE:理论驱动型社会智能评测基准 (2026-06-12)
- 25个Skills详解:从生产力清单到AI工作流资产
- OpenAI科学家Noam Brown:AI的真正上限,可能根本没人测得起
- 开源 AI 知识管理搭档 Obsidian + Claude Code 完整集成指南 (2026-05-08)
- 我又把 Obsidian 知识库升级了:现在它能自己长出知识网络
- obsidian-llm-wiki-local: Obsidian本地AI知识图谱自动构建工具
- 对比离线蒸馏!为什么选择OPD?
- OneReason:当推荐系统学会思考
- 800行代码实现 Open Claw 的 Tool、消息总线、子Agent管理架构 (2026-05-09)
- Open Notebook:开源版 Google Notebook LM 替代品
- 明星开源项目,为什么开始离开 GitHub?
- OpenTelemetry eBPF Instrumentation (OBI) 零代码全栈可观测性 — 阿里云云原生深度解读
- Codex 5.21 更新:AI 编程助手开始变成电脑工作代理
- openai-codex-jasonliu-maxxing-playbook
- openai gpt realtime voice models qbitai
- OpenAI Realtime API 架构首次公开 (2026-05-07)
- 一文看懂 OpenAI 开源的 Codex 编排规范:Symphony
- OpenAI三个语音模型发布同传被杀死 (2026-05-07)
- OpenAI发布新一代实时语音模型,能够像人说话一样进行推理、翻译和转录 (2026-05-16)
- openai官方cli上线跟复杂的sdk说拜拜
- openai携手五巨头开源革命性超算协议一举解决超大集群llm训练不稳定和网络性能难题
- OpenChronicle:把AI记忆变成可复用的基础设施 (2026-04-25)
- 你的 AI Agent 真的在受控运行吗? (2026-04-28)
- 花费 2 个星期写了 8 篇 OpenClaw 源码拆解文章,我发现90% 的人对龙虾的理解都太表面了,深层次的真相竟然是这个
- 800行代码实现 Open Claw 的 Tool、消息总线、子Agent管理架构 (2026-04-24)
- 别再亲自写Prompt了!OpenClaw与Claude Code创始人已经用'循环'让一堆Agent自动打工了
- 为OpenClaw配置网盘空间的最佳实践
- openclaw comprehensive guide 32k chars
- OpenClaw与Hermes:源码里的 AI Agent 架构知识大复盘
- openclaw multi agent team practice v2
- openclaw multi agent team practice v2
- openclaw-prompt-context-harness (2026-04-25)
- openclaw-完全指南这可能是全网最新最全的系统化教程了32w字建议收藏-v2
- 1.6万 Star,AI Agent 赛道又杀出一匹黑马!
- openspec-spec-driven-development-trae-solo
- OpenSpec + Superpowers 融合方案下线记:2 周 3 次实测 + 3 个测试识别缝合怪 + 加法 vs 减法传播学 + Plan Mode + Superpowers + ASD 最终方案
- 面试官问:'你都是Vibe Coding,怎么保证质量?'——用OpenSpec、Superpowers、gstack把Vibe Coding 拉回到工程交付 (2026-06-12)
- OpenSpec 四步法深度复盘:流程完整 ≠ 代码正确
- 刚刚Opus 4.7发布,相比4.6核心变化,与Claude Code搭配最佳实践 (2026-05-09)
- PersonaVLM: Long-term Personalized Multimodal LLM (2026-04-27)
- phoneworld-mobile-agent-scaling-mock-environments-tencent-hunyuan-arxiv-2605-29486
- Coding Harness 工程本质:从 Pi 到 OpenClaw (2026-05-07)
- pilotdeck-agent-os-openbmb-tsinghua
- pilotdeck-data派thu-2026
- pith-train-agent-native-moe
- pixelle-video-aidc-ali-international-2026
- AI第一次科研竞赛中击败人类!Opus 4.7狂飙2930步创世界纪录
- 如何构建生产准备的AI代理:MCP、CLI与技能——适合合适的工作的工具
- 生产级 Harness 的 12 大组件以及主流框架对比 (2026-05-12)
- prompt-debugger-compare-templates-winty (2026-05-12)
- 知识库问答 @文档:从 DOM 方案到 ProseMirror 落地(vivo 互联网项目团队) (2026-06-04)
- protocol-h-hierarchical-agentic-rag-enterprise (2026-05-07)
- Qoder 1.0正式发布!从AI IDE迈向智能体自主开发工作台
- Agent Skills 终于有 UI 了 (2026-04-27)
- qoder-skills-完全指南从零开始让-ai-按你的标准执行-v2
- Qoder Skills 完全指南:从零开始,让 AI 按你的标准执行 (2026-05-16)
- Qoder 发布团队知识引擎:组织级知识记忆是 Harness 自进化的重要组件
- QQ音乐 Harness Engineering 实践(大仓多服务场景)
- Subagents 详解:Claude Code 如何避免上下文污染
- R1复现:拒绝采样微调加速RL收敛及模型遗忘问题探究 (2026-05-11)
- rag chunk embedding rerank pipeline
- RAG深度解析:分块、向量化、召回、重排,才是“蒸馏同事skill”的关键 (2026-05-10)
- RAG 全链路技术详解
- rag-vector-knowledge-graph-ontology (2026-05-07)
- RAG vs LLM Wiki:企业知识库架构的技术深剖 (2026-05-07)
- Ralph Loop 不够用:长时间 Agent 还缺这 3 件事 (2026-05-14)
- 复杂业务场景下 RCA Agent 的探索实践
- ReAct、Plan-and-Execute、Reflection、Multi-Agent 模式选型指南:5 步决策树 + AgentScope 完整 Demo
- Recent Developments in LLM Architectures: KV Sharing, mHC, and Compressed Attention(机器之心中文全文)
- 100 年压缩到 100 天!红杉资本:这就是 AGI (2026-05-10)
- Redis之父下场,给DeepSeek V4单独造了一台推理引擎 (2026-05-10)
- Rein:4 模块 + 5 类型边界防止 agent.go 膨胀到 3000 行 (2026-06-04)
- Review Agent:后台复盘 Agent 如何判断什么值得保存
- rocket-mq-5-litetopic-ai-agent-async-cloudnative
- Apache RocketMQ 5.5.0 开源 LiteTopic:百万级 AI 会话专属通道
- Harness 工程可视化:Vibe Coding 中重建工程可控性 (2026-05-07)
- 开源的多Agent协同交付平台Routa
- 用好 Claude 的 18 个动作:搭一个个人 AI 工作台
- saas-bench-gui-agent-eval-unipat
- SAG 知识引擎:用 SQL 做 RAG — VibeCoder 源码级解读 (MCP 集成 + 检索算法 + 消融实验) (2026-06-17)
- sag-sql-retrieval-augmented-generation-zleap-ai-2026-06-16
- Salesforce 主动砍掉了界面,a16z 合伙人说:护城河从来不在那里
- GUI Agent「记与学」双修,长程任务有了专属记忆增强型自进化框架
- Self-Evolving Agents 系统性综述(厦门大学等多机构联合) (2026-05-07)
- 上海AI Lab 提出 Self-Harness:Agent自己改Harness,通过率最高涨138% (2026-06-12)
- 让大模型学会「自己教自己」!京东&中科院信工所连发三篇论文定义Self-TaughtRLVR
- 商汤开源 SenseNova-U1:一个模型,同时「看懂」和「画懂」 (2026-04-29)
- 第一步:安装 Remotion Best Practice Skill (2026-04-30)
- AI 总跑偏?Skill 太人话了:企业级 Skill 8 块最小骨架 + 8 条 checklist
- 重新定义Skill开发:保姆级教程&一站式开发助手发布 (2026-05-18)
- 当我把AI变成一个"算法":Skill工程化设计的心路历程
- Skill Factory:三天手搓面向Harness设计的技能工厂
- 10篇论文看懂AI Agent Skill:表示、执行、评估与进化 (2026-05-07)
- AI 越用越聪明公司却越用越白用:Skill Hub 是企业级 AI 经验资产化的关键
- AI Skill 也会"退休",企业到底怎么管?— winty Skill 生命周期 6 阶段治理(创建/审核/发布/灰度/运行/废弃 + 数据化迁移规则 + Dashboard + 真实遗忘事故复盘) (2026-06-17)
- skill-product-philosophy-guicang-爆款经验-2026-06-12
- Skill 想进生产?先闯这 4 道死关
- Skill Retrieval Augmentation for Agentic AI — 清华 Skill-RAG (2026-04-29)
- Skill自进化三路线:Trace2Skill归纳法 / EvoSkill验证闭环 / SkillOpt训练范式
- AI Agent 架构设计(七):Skills 系统设计(OpenClaw、Claude Code、Hermes Agent 对比) (2026-04-27)
- Skill 升了版,你说不清楚哪里变好了吧?
- Skill 改一版烂一版?少了版本号谁都救不了
- 工作流的 Skill 怎么写?从 7 个顶级 Skill 中提炼的模式与最佳实践 (2026-04-27)
- 阿里SkillClaw:让 Agent 技能在真实使用中集体进化 (2026-04-25)
- 高德 SkillClaw:让 Agent Skill 学会进化——跨会话、跨Agent、跨设备、跨用户 (2026-04-24)
- Agent 技能夜间自进化——阿里开源 SkillClaw,最高提升 88% (2026-04-11)
- SkillClaw × Nacos:从一次 Agent 会话到可治理 Skill Registry 的自动演化闭环
- 阿里SkillClaw:让 Agent 技能在真实使用中集体进化
- skillopt-microsoft-train-skill-hyman's-blog
- 别再手写 Skill 了!微软最新研究:像神经网络一样训练 Skill
- SkillOpt:把 Agent 技能文档变成可训练对象 — 微软+上海交大等团队 (像训练小程序一样训练 agent 技能) (2026-06-04)
- 面向Skills编程-淘宝企业购端对端研发提效实践
- Skills Registry 公测开启:为企业打造私有的 Skill 管理中心 (2026-05-09)
- SkillSieve:Agent Skill 安全检测三层框架 (2026-05-01)
- SkillTrustBench:首个 Agent 技能安全评测基准 (港中文深圳 + 腾讯朱雀实验室) (2026-06-16)
- 浙大开源SkillX:全自动构建Agent技能知识库,即插即用提升10%性能 (2026-05-01)
- SkillX: Automatically Constructing Skill Knowledge Bases for Agents (2026-04-27)
- Agent自我进化为什么这么难?small Hermes实践总结
- Snowflake 迈向 Agentic Enterprise 的关键一跃 — Summit 2026 Platform Keynote 解读
- Hermes Agent Eval Harness 深度解析:从执行轨迹到可验证的 Skill 进化(7 模块闭环 + 6 类 Verifier + 3 类 Regression + 6 阶段落地路径)
- Spec as AIOS:AI-Native 全栈交付的抗熵架构
- AI实践|基于 Spring AI 从0到1构建 AI Agent (2026-04-25)
- ssd-spec-driven-development-harness-asd-shuge-2026-06-17 (2026-06-17)
- AI短漫剧的Netflix来了!恒星AI推出全球首个影视级AI Agent「Starfilm」,打造AI短漫剧制作的“超级梦工厂”
- State of Memory in Agent Harness — mem0 视角的九大 harness 横评
- 大模型可控新突破:Steering 机制、评估体系与开源落地
- 龙虾之父教你省钱:开源Skill给你的Skill减肥
- 他的 Agent 昨晚替他把公司运转了一遍,你的早会才刚开始
- sub-agent-vs-agent-team-selection-guide (2026-04-30)
- subagents-详解claude-code-如何避免上下文污染-v2
- Subagents 详解:Claude Code 如何避免上下文污染
- Superpowers 6.0 源码级深度分析:反作弊重写——reviewer 只读怀疑论者 + 上下文经济学 + progress ledger + model 纪律
- Superpowers 6.0 深度解读:SDD 评审重写 + 文件交接 + 多平台支持
- Superpowers:给 Claude Code 装上'工程大脑'
- 给 Claude Code 装上「方法论」:深入解读 superpowers
- 别再给智能体硬塞 System Prompt 了
- 花费 2 个星期写了 8 篇 OpenClaw 源码拆解文章,我发现90% 的人对龙虾的理解都太表面了,深层次的真相竟然是这个
- 大淘宝数字SRE:AI主导的代码质量治理实践
- 万级实时推理的商品领域Agent实践思考和总结
- tencent-ai-infra-backend-engineer-huangrunpeng (2026-05-08)
- Harness不是目的,知识才是护城河 —— 一个AI工程交付团队的知识沉淀实践
- Harness Engineering:AI 能在真正"出事会炸"的后端系统里写代码吗? (2026-04-28)
- 腾讯混元Hy3-preview发布 (2026-04-23)
- 腾讯混元新里程碑:Hy3 preview 发布开源,Agent 表现全面提升
- Harness不是目的,知识才是护城河 —— 一个AI工程交付团队的知识沉淀实践 (2026-04-27)
- AI Agent & Skill 测评方案及落地实践 — 腾讯 TEG 网关测试团队 (2026-06-16)
- 让AI成为真正的社会生产力——跨越Token效率门槛走向AI普惠
- tencent vibe coding to agentic engineering backend
- 腾讯云Agent Memory节省61% Token提升52%成功率的诀窍:Mermaid无限画布×上下文卸载
- Thin Harness, Fat Skills:AI工程架构的本质 (2026-04-25)
- thinking-machines-interaction-models-ai-cold (2026-05-13)
- 从零设计一个准生产级 LLM Agent:可靠性、上下文与恢复机制
- 一文看懂三种 RAG 架构:Classic RAG、Graph RAG 与 Agentic RAG (2026-05-16)
- AI 编程总失控?Comet + OpenSpec + Superpowers 用文件系统管住 AI — 术哥源码级深度分析(三器流水线:WHAT/HOW/WHEN/NEXT + 9 平台 Hook + 上下文压缩 + 多 Agent 协作 + Shell+YAML 工程取舍)
- 三器合一:gstack + Superpowers + OpenSpec 工程化 AI 编程实战 (2026-05-12)
- 时间序列预测增强方法总结:频域、分解、patch
- 淘天营销中后台生码工作流最佳实践 (2026-04-27)
- 从定制Workflow到AI自主决策的架构演进——以TMIC AI小新为例
- 一篇搞懂 AI Coding Agent 的 Token 成本控制
- 让AI成为真正的社会生产力——跨越Token效率门槛走向AI普惠
- token级,精准控制生成长度:3B模型击败GPT 5.4、Claude (2026-05-10)
- topological-trouble-transformers-state-tracking-deepmind-2026-06-17 (2026-06-17)
- Trace2Skill 把"轨迹里的局部经验"蒸馏成可迁移的 Agent Skills
- trae-solo-work-feishu-bitable-tutorial
- trajectory-balance-asynchrony-tba-bengio-papweekly
- Triton L2缓存命中优化矩阵乘法(fp16&int8)详解及性能测试 (2026-05-09)
- 细思极恐!Agent暗藏风险,清华团队打出组合拳,全链路一网打尽 (2026-05-07)
- 清华大学:驾驭工程 (Harness Engineering) 研究报告 (2026-04-10)
- tsinghua-self-evolving-skill-agent
- 一文读懂 12 个 Agent 工程设计底层逻辑
- 两篇Harness论文:微软M⋆记忆进化 + 谷歌AutoHarness动作约束 (2026-04-25)
- understand-anything-code-knowledge-graph-lum-jike
- Understand Anything:把代码库变成可以探索的知识图谱 (2026-06-11)
- verify-data:一个端到端的数据验数 Agent Skill
- 7个月,234次提交,1690行代码:AI编程大型翻车现场:我决定全部作废,手动重写! (2026-05-13)
- video-rag-chunking-strategy-deephub-imba
- 搞定视觉幻觉难题,这套并行推理框架设计很巧妙 (2026-06-10)
- AI 导购在 vivo 官网的落地实践
- 墙比模型更重要:Stripe Minions + 字节 DeerFlow 2.0 + 蚂蚁支小助 三家公司的同结论 (2026-06-01)
- wall-oss-05-pretraining-embodied-ai-x-square-robot
- 王云鹤眼中的Harness:复杂优化问题,AGI灵魂争夺之战
- 微软 1000 行代码,把 Claude Opus 干翻了 15 分
- 蔚蓝BabyAlpha A3消费级机器狗撕开英伟达垄断
- 微信读书出官方skill了,但它还差关键一步
- workbuddy-llm-wiki-obsidian-practice
- wow-harness v3 —— AI 开发的治理协议:跨 session 一致性 + Agent 组织化 (2026-06-04)
- xai-grok-musk-training-new-model-wechat
- xai解散但Grok还活着 (2026-05-08)
- Xiaomi Dasheng — 通用声音基座模型(8卡起步的 AI 工程实践)
- OpenAI Tax AI 自我进化:Codex Eval 闭环 6 周 25%→86% + Symphony + MOSS
- xopd-on-policy-distillation-landscape-banana-2026
- Yann LeCun谈LLM不是智能与世界模型JEPA
- 200美元搞定400万美元的项目!YC总裁:同时调用15个Agent,每月交付数十万行代码,Claude Code和Codex协作开发,用Token换时间亿万富翁
- 如何用AI打造一家自我进化的公司
- 我把 YC 最新一批 196 家公司全扒了一遍——AI 已经不再是差异点了
- yidian tianxia context engineering agentic ai qcon
- 智子芯元:用大模型+运筹优化打造计算加速智能体,两轮累计融资近亿
- 一文带你弄懂-ai-圈爆火的新概念harness-engineering-v2
- 一次构建,随处复用:Python 中的泛型仓库模式 (2026-05-10)
- 一次构建随处复用python-泛型仓库模式
- 不改模型、不降质量,谷歌让Gemma 4快了3倍:本地跑大模型彻底变天 (2026-05-16)
- 不用再学ai了生成结果包稳的agent来了
- 不用学AI了!圈内公开的秘密:顶级玩家已开始让AI用AI
- 五一活动来啦 (2026-05-16)
- 今天起,Claude正式接入Office全家桶,跨应用还能共享记忆 (2026-05-16)
- 从-30-分钟手搓-agent到-harness-成为新后端
- 从 Anthropic 到 Google:Agent Skills 正在进入“设计模式”阶段 (2026-05-10)
- 从 CPU 到 GPU 全链路可信,百度智能云新一代 AI 机密计算实例的探索与落地 (2026-05-16)
- 从全量启动到最小核: 手淘外链唤端链路的三次架构演进 (2026-05-16)
- 从全量启动到最小核手淘外链唤端链路的三次架构演进
- 从多智能体编排到AI自主决策:资损防控体系的架构演进
- 从提需求到部署发布,全AI全自动化后,研发效能全面跃升 (2026-05-16)
- 企业级多-agent-规模化落地怎么做群虾智能-ai-沙龙-ppt-限时领取
- 企微的这些新功能,补齐了AI在你公司的最后一公里
- 估值3000亿63家新实验室杀疯了murati贝佐斯集体押注下一代ai
- 你不知道的-agent原理架构与工程实践-v2
- 你写的 Skill,及格了吗? (2026-05-10)
- 你的一人公司品牌部,带着Image-2模型的lovart中文版来了 (2026-05-11)
- 使用 Agent Skills 做知识库检索,能比传统 RAG 效果更好吗? (2026-05-14)
- 使用Claude Code:session管理与1M上下文 (2026-05-16)
- 全球AI新王诞生!Anthropic估值冲爆1.2万亿,首次反超OpenAI (2026-05-16)
- 准备开一个新坑从零复刻一个-claude-codenn目标是在这个过程中和大家一起学习-claude-code-的-harness-是如何做的nnclaude-
- 刚刚,OpenAI 放出三个语音模型,顺便杀死了「同传」
- 刚刚opus-47发布相比46核心变化与claude-code搭配最佳实践-v2
- 刚刚Opus 4.7发布,相比4.6核心变化,与Claude Code搭配最佳实践 (2026-05-16)
- 别为了用龙虾而用龙虾一个技术管理者折腾三周唯一留下的场景却是这个
- 别再把上下文当聊天记录 (2026-05-10)
- 十年老技术开发的 AI Agent 探索之路 (2026-05-16)
- 升职那天我才明白自己是来填坑的
- 卡片式对话的协议方案探索和思考 (2026-05-09)
- 向量库是rag的前菜知识图谱是答案本体论是灵魂-v2
- 向量库是RAG的前菜,知识图谱是答案,本体论是灵魂 (2026-05-16)
- 吴恩达2026新课上线!3小时包教包会,零代码小白也能成为AI超级玩家 (2026-05-16)
- 吴恩达:AI 将最先杀死前端 (2026-05-10)
- 告别“氛围编程”:基于 Harness 治理和 SDD 的团队级 AI 研发范式演进与实践 (2026-05-10)
- 国产顶尖模型 benchmark 评分那么高,可实际效果为什么差?看完 Anthropic 这篇博客,刷分的因素太单一了 (2026-05-10)
- 在 RDS PostgreSQL 中实现 RaBitQ 量化 (2026-05-14)
- 大反转!马斯克牵手对手 Dario,Anthropic 与 SpaceX 罕见合作 (2026-05-16)
- 天猫新品团队ai编码实战指南下-v2
- 天猫新品团队AI编码实战指南(下) (2026-05-16)
- 天猫新品营销技术团队ai编码实战指南上-v2
- 奥特曼最险一战:前女CTO当庭翻脸,OpenAI权斗彻底打到台前
- 如何谈合作、找工作 (2026-05-15)
- 实践教程真实ai客服落地全流程意图识别混合检索到数据飞轮-v2
- 【实践教程】真实AI客服落地全流程:意图识别、混合检索到数据飞轮 (2026-05-16)
- 工作流的 Skill 怎么写?从 7 个顶级 Skill 中提炼的模式与最佳实践 (2026-05-10)
- 开源-ai-知识管理搭档-obsidian-claude-code-完整集成指南-v2
- 开源 AI 知识管理搭档 Obsidian + Claude Code 完整集成指南
- 强化学习如何不用奖励模型提高通用问题推理能力 (2026-05-11)
- 很多企业做完 AI PoC,为什么还是上不了生产 (2026-05-16)
- 快手首个打工人Agent来了!工作秒变桌面软件:零代码、不烧token (2026-05-11)
- 快来,和AI实战派一起AI!AIGC峰会最新嘉宾阵容来了
- 我把 Claude Design 做成了 Skill,人人都能成为顶级网站设计师 (2026-05-09)
- 我把 Karpathy 的 AutoResearch 搬到了软件开发领域,效果炸了 (2026-05-10)
- 我把mac留在家用手机让trae-solo替我打了一天工
- 我用 SKILL.md 做了一个简历生成器 (2026-05-09)
- 我给Hermes配了4个Agent,真正有用的是这些事 (2026-05-11)
- 打造可靠的-ai-编程环境claude-code-hooks-完整开发者指南-v2
- 打造可靠的 AI 编程环境:Claude Code Hooks 完整开发者指南 (2026-05-16)
- 提速4.48倍!哈工大华为新框架让扩散大模型精度无损、推理起飞 (2026-05-16)
- 无需复杂环境搭建教你用自己的-agent-玩转-moltbook
- 日志别再print了!深入对比Python三大日志方案 (2026-05-10)
- 是时候讨论“人类配额制”了:应对AI就业冲击的一种可能性探讨 (2026-05-09)
- 构建基于多智能体架构的深度思考交易系统-v2
- 构建基于多智能体架构的深度思考交易系统 (2026-05-16)
- 民生银行基于规格驱动开发(SDD)的 CodeAgent 私域研发探索与实践 (2026-05-10)
- 注定改变历史的一代人 (2026-05-10)
- 淘天营销中后台生码工作流最佳实践 (2026-05-09)
- 深入理解-claude-code-源码中的-agent-harness-构建之道-v2
- 深势科技携手阿里云 AgentRun,加速科研 AI Agent 全速运行 (2026-05-16)
- 特斯拉百万年薪招数据标注员,朝九晚五,无需AI经验 (2026-05-11)
- 直播预约 | 数据引擎:具身智能的下一个决胜局 (2026-05-16)
- 突发!Anthropic拿下马斯克Colossus 1全部算力:Claude要放开用了 (2026-05-16)
- 精选 10 个开发者常用的 AI 智能体技能(Agent Skills) (2026-05-10)
- 精选 8 个 UI 设计师必备的 AI 智能体技能(Agent Skills) (2026-05-09)
- 给“氛围编程”系上安全带:阿里集团 AI 代码评审实践与 Benchmark 开源
- 腾讯员工公寓曝光,竟是这样的布置!
- 腾讯混元新里程碑:Hy3 preview 发布开源,Agent 表现全面提升 (2026-05-16)
- 腾讯研究院AI速递 20260430
- 腾讯研究院AI速递 20260506 (2026-05-16)
- 腾讯研究院ai速递-20260507
- 腾讯研究院AI速递 20260508 (2026-05-14)
- 要实现一个工作流选择-agent-skills-还是-ai-表格
- 让AI成为真正的社会生产力——跨越Token效率门槛走向AI普惠
- 语音输入喊了这么多年,千问电脑版一出手就把键盘卷没了? (2026-05-11)
- 读完 Claude Code 和 OpenClaw 的 memory 源码,我对"Agent记忆需要向量数据库"这件事产生了怀疑…… (2026-05-16)
- 读完这篇你就搞懂-deepseek-v4-了-v2
- 读完这篇,你就搞懂 DeepSeek v4 了 (2026-05-16)
- 豆包 Seed 2.0 Lite升级:给 Agent 装上眼睛和耳朵 (2026-05-16)
- 还在手写-osgetenvpydantic-settings-让你配置管理效率翻倍
- 这张信息图居然是8b开源模型做的
- 这样的程序员,应该招吗? (2026-05-16)
- 长周期-agent-详解-从-ralph-loop-到可接管-harness (2026-05-10)
- 阿里云可观测 2026 年 4 月产品动态 (2026-05-16)
- 阿里巴巴&蚂蚁 LoongSuite GenAI 可观测语义规范:从统一数据语言到规模化落地
- 面向电商直播场景的全模态大模型推理加速方案 (2026-05-09)
- 鹅厂员工,最近看的一本书是什么? (2026-05-09)
- 龙虾装上了可以用来干啥分享下我的-openclaw-多智能体团队搭建经验-v2
aws.amazon.com(239 篇)¶
- 750b-moe-model-roce-cluster-migration-aws-efa-prefill-decode-disaggregation
- accelerate-campaign-workflow-with-insights-from-adobe-market
- accelerate-llm-model-loading-and-increase-context-windows-wi
- aderant-transforms-cloud-operations-with-amazon-quick
- Agentic AI基础设施实践经验系列(九):Context Engineering 上下文工程 | 亚马逊AWS官方博客 (2026-02-24)
- 让 AI 代理自己付钱:基于 Amazon Bedrock AgentCore 与 x402 的 Agentic Payment
- 基于Strands和AgentCore 实现Agentic Scheduler 在多Region自动编排推理GPU算力 | 亚马逊AWS官方博客 (2026-02-25)
- agentops-operationalize-agentic-ai-at-scale-with-amazon-bedr
- ai agent 的迁移与现代化 使用 amazon bedrock agentcore 将 openclaw 从单机改造为多租户 serverless 架构
- AI 驱动的 Graviton 迁移评估:Kiro Power 实战指南 | 亚马逊AWS官方博客 (2026-03-27)
- AI 驱动的跨云网络搭建:用 Claude Code 和 Kiro CLI 实现 AWS-腾讯云 IPSec VPN 双隧道互联 | 亚马逊AWS官方博客 (2026-04-24)
- ai-powered-bi-with-snowflake-and-amazon-quick
- ai-research-assistant-from-idea-to-app
- 让 AI 理解你的组件库:新一代智能 D2C架构 — 基于 AWS Kiro MCP Skills 的智能转换实践 | 亚马逊AWS官方博客 (2025-12-08)
- Amazon Bedrock AgentCore 为部署可信人工智能代理增加了质量评估和策略控制 | 亚马逊AWS官方博客 (2026-02-02)
- Dify集成Amazon Bedrock AgentCore Browser 实现更强大的信息获取和分析能力 | 亚马逊AWS官方博客 (2026-02-02)
- amazon-bedrock-agentcore-harness-is-now-generally-available-
- 快时尚电商行业智能体设计思路与应用实践(七)Amazon Bedrock AgentCore Runtime 深度解析和场景分析 | 亚马逊AWS官方博客 (2026-01-30)
- 别让你的 Amazon Bedrock 模型为他人打工——API 调用安全防护指南
- amazon-bedrock-claude-prompt-cache-strategy
- Amazon CloudFront部署小指南(二十四):将CloudFront “多域名”改造为”多租户”架构 | 亚马逊AWS官方博客 (2026-04-07)
- amazon-ecs-引入新的高分辨率指标以实现服务快速自动扩缩
- 基于 Amazon IoT Core 与 Kiro 构建可迁移的工业 IoT 数据管道
- amazon-nova-act-is-now-hipaa-eligible
- Amazon Nova Lite Fine-Tuning: 高性价比的视觉检测模型微调案例与实践 | 亚马逊AWS官方博客 (2026-01-07)
- Manufacturing Intelligence with Amazon Nova Multimodal Embeddings
- amazon-quick-accelerating-the-path-from-enterprise-data-to-ai-powered-decisions
- amazon-quick-arns-cross-account-migration-and-namespace-perm
- 用 Amazon Quick + Bedrock AgentCore 打造对话式 FinOps 助手
- amazon-quick-mcp-kdbx-time-series
- Amazon Redshift 推出带有集成数据湖查询引擎的基于 AWS Graviton 的 RG 实例
- amazon-s3-annotations-queryable-context
- amazon-sagemaker-ai-async-inference-now-supports-inline-requ
- 用 Amazon SageMaker AI 与 Qualcomm AI Hub 打通从云端训练到端侧 NPU 的交付闭环
- amazon-workspaces-applications-quick-build
- Announcing AWS CDK Mixins: Composable Abstractions for AWS Resources | Amazon Web Services
- Announcing OpenAI-compatible API support for Amazon SageMaker AI Endpoints
- anthropic-claude-fable-5-on-aws内置保护措施的-mythos-级功能现已推出
- Automate Schema Generation for Intelligent Document Processing
- aws bedrock agentcore quality optimization flywheel
- Halliburton enhances seismic workflow creation with Amazon Bedrock and Generative AI
- aws bedrock intelligence message defense
- aws-bedrock-ops-alert
- aws bedrock serverless async inference sqs lambda
- 利用 AWS Budget 实现 Amazon Bedrock 用量监控、超预算告警与自动中断方案
- aws-china-enterprise-intelligent-validation-prototype-to-production
- aws-china-enterprise-intelligent-why-evaluation
- aws-china-how-to-build-enterprise-intelligent-agent
- aws-cn-intelligent-device-assistant-consumer-agent-2026
- Introducing AWS Continuum for security at machine speed
- Production-Ready Autonomous Incident Resolution with AWS DevOps Agent (now GA) and Datadog MCP Server | Amazon Web Services
- aws-devops-agent-mcp-china-part2-multi-account-roles-anywhere
- AWS DevOps Agent 接入 AWS 中国区(一):Partition 隔离与 MCP 单账号桥接
- AWS DevOps Agent × MCP Server:打通混合云网络排障的最后一公里
- AWS DevOps Agent 实战:云网络故障自主调查与修复建议
- aws ec2 capacity blocks gpu ml
- aws hapag lloyd bedrock customer feedback
- aws mlflow v310 generative ai development
- aws quicksight dataset qa natural language
- aws quicksight dataset qa tara case
- aws-sagemaker-azerbaijani-lm
- aws sagemaker capacity aware inference fallback
- aws-sagemaker-sft-dpo-tool-calling
- aws-software-supply-chain-security-well-architected
- aws transform ezconvertbi bi migration
- aws-waf-ai-traffic-monetization-bot-content-access
- AWS 一周综述:Amazon Bedrock AgentCore 付款、适用于 AWS 的 Agent 工具套件等(2026 年 5 月 11 日)
- aws-一周综述aws-transform-上线一周年aws-云端-claude-platformec2-m3-ultr
- Back up and restore your Amazon EKS cluster resources using Velero | Amazon Web Services
- based-on-prowler-genai-build-fintech-intelligent-compliance-2
- bedrock-agentcore-secrets-manager-identity
- bedrock-claude-litellm-websearch-interception-配置指南
- 对图像内容进行精确分析 — Bedrock 多模态案例实践
- better-decisions-at-scale-how-mathematical-optimization-deli
- 使用 Kiro AI IDE 开发 AWS CDK 部署架构:从模糊需求到三层堆栈的协作实战 | 亚马逊AWS官方博客 (2026-02-02)
- break-the-context-window-barrier-with-amazon-bedrock-agentco
- build-a-healthcare-appointment-agent-with-amazon-nova-2-soni
- build-a-meeting-prep-and-follow-up-assistant-with-amazon-quick-and-cisco-webex-mcp-servers
- build-ai-agents-for-business-intelligence-with-amazon-bedroc
- build-ai-powered-dashboard-automation-agents-with-nlp-on-ama
- build-an-agentic-incident-triage-assistant-with-amazon-quick-and-new-relic
- build-an-ai-powered-equipment-repair-assistant-using-amazon-
- build-an-enterprise-observability-solution-for-amazon-quick
- build-context-rich-research-agents-with-deep-agents-and-bedr
- build-custom-code-based-evaluators-in-amazon-bedrock-agentco
- Build financial document processing with Pulse AI and Amazon Bedrock
- 基于 Amazon EKS 和 Graviton 构建多租户 AI Agent 平台:OpenClaw on Kubernetes 实践 | 亚马逊AWS官方博客 (2026-03-20)
- build-real-time-voice-applications-with-amazon-sagemaker-ai
- Build real-time voice streaming applications with Amazon Nova Sonic and WebRTC
- build-self-service-aws-health-analytics-to-find-actionable-h
- building-a-secure-auth-code-flow-setup-using-agentcore-gatew
- building-agentic-ai-applications-with-a-modern-data-mesh-str
- building-ai-agents-for-business-support-using-amazon-bedrock
- 用 Kiro构建 AI:基于 AWS 基础设施快速构建企业级 Agentic AI 平台 | 亚马逊AWS官方博客 (2026-03-04)
- 基于Bedrock AgentCore+Strands构建企业级智能搜索平台实践 | 亚马逊AWS官方博客 (2026-02-24)
- building-multi-tenant-agents-with-amazon-bedrock-agentcore
- building-pay-per-intelligence-for-ai-agents-how-ampersend-us
- building-supercharger-how-rocket-close-optimized-title-opera
- building-web-search-enabled-agents-with-strands-and-exa
- CI&T基于 Amazon Bedrock AgentCore 与 OpenClaw 的企业级智能运维最佳实践 | 亚马逊AWS官方博客 (2026-04-17)
- claude-code-aws-bedrock-guide
- Claude Code 接入自建开源模型:企业私有化与降本实践 | 亚马逊AWS官方博客 (2026-04-02)
- The New Era of Cloud AI Mobile Testing: Amazon Device Farm MCP Server Practical Guide | 亚马逊AWS官方博客 (2026-02-02)
- control-where-your-ai-agents-can-browse-with-chrome-enterprise-policies-on-amazo
- Cost effective deployment of vision-language models for pet behavior detection on AWS Inferentia2
- data-for-ai明其所耗知其所因让每一分-token-消耗都可量化的全栈实践
- 快时尚电商行业智能体设计思路与应用实践(六)借助 Amazon Bedrock AgentCore MCP Server,Amazon Bedrock,Strands Agents,Kiro 实现智能体极速研发 | 亚马逊AWS官方博客 (2026-02-02)
- 使用 Kiro AI IDE 开发 基于Amazon EMR 的Flink 智能监控系统实践 | 亚马逊AWS官方博客 (2026-02-02)
- direct-connect-dx-迁移最佳实践
- 快时尚电商行业智能体设计思路与应用实践(五)借助 AgentCore Runtime 与 Bedrock 模型平台,轻松实现 Claude Agent SDK 的生产级部署 | 亚马逊AWS官方博客 (2026-02-02)
- ec2-nat-instance-deploy-practice-aws-china
- embed-the-world-multimodal-ai-for-searchable-aerial-imagery
- 让 Kiro 和 Claude Code 响应 IM 消息:用 ACP Bridge 打造异步 AI 编程工作流 | 亚马逊AWS官方博客 (2026-03-27)
- enable-safe-agentic-payments-with-built-in-guardrails-using-
- end-to-end-encrypted-ml-inference-with-amazon-sagemaker-ai-a
- 基于Strands SDK 构建的企业智能问数解决方案实践 | 亚马逊AWS官方博客 (2026-02-24)
- 企业级OpenClaw安全部署架构指南 | 亚马逊AWS官方博客 (2026-04-23)
- evaluate-ai-agents-systematically-with-agent-evalkit
- evaluate-your-amazon-nova-sonic-voice-agent-at-scale-no-micr
- OpenClaw 在电商平台的应用场景探索 | 亚马逊AWS官方博客 (2026-03-11)
- extending-mcp-support-for-amazon-bedrock-agentcore-gateway
- Extract Data with On-demand and Batch Pipelines Dynamically
- 快时尚电商行业智能体设计思路与应用实践(八)基于 WebSocket 的语音系统:Nova 2 Sonic, AgentCore, Strands Agents 企业级架构实践 | 亚马逊AWS官方博客 (2026-01-30)
- Fine-tune LLM with Databricks Unity Catalog and Amazon SageMaker AI
- 从代码到分子系列:一场由 AI 驱动的 EGFR 抑制剂发现之旅 — 深度融合 AWS Bedrock与 Claude Code/Claude Agent Skills,生命健康行业的科学活动探微 | 亚马逊AWS官方博客 (2026-02-09)
- 从手动到智能:用 Kiro CLI + OpenSearch MCP 让每个人都成为 OpenSearch 专家 | 亚马逊AWS官方博客 (2026-01-30)
- from-pdfs-to-insights-architecting-an-intelligent-document-p
- from-siloed-data-to-unified-insights-cross-account-athena-access-for-amazon-quic
- Fundamental’s Large Tabular Model NEXUS is now available on Amazon SageMaker JumpStart
- 基于 MIG 技术在 Amazon SageMaker HyperPod 上实现 GPU 虚拟化的最佳实践 | 亚马逊AWS官方博客 (2026-02-24)
- Habby 游戏借助 AWS DevOps Agent 实现智能运维最佳实践
- hands-free-first-notice-of-loss-using-strands-agents-and-ama
- how-amazon-finance-streamlines-regulatory-inquiries-by-using
- how-baz-improved-its-ai-agent-code-review-accuracy-using-ama
- how-frontier-teams-are-reinventing-ai-native-development
- how-loka-built-a-natural-low-latency-voice-agent-with-amazon
- implementing-super-resolution-by-deploying-seedvr2-on-amazon
- improve-bot-accuracy-with-amazon-lex-assisted-nlu
- integrate-atlassian-confluence-cloud-with-amazon-quick
- integrating-aws-api-mcp-server-with-amazon-quick-using-amazo
- 智能体驱动测试变革:让智能体成为测试第一性 之三 用 Web Bot Auth 为 AgentCore Browser Tool 打造可信身份 | 亚马逊AWS官方博客 (2026-02-02)
- 基于Bedrock Agentcore 实现智能成本分析与告警系统 | 亚马逊AWS官方博客 (2026-03-04)
- Intelligent radiology workflow optimization with AI agents
- Introducing Claude Platform on AWS: Anthropic's native platform, through your AWS account
- introducing-gemma-4-models-on-amazon-bedrock
- Introducing OS Level Actions in Amazon Bedrock AgentCore Browser
- introducing-web-search-on-amazon-bedrock-agentcore
- its-safe-to-close-your-laptop-now-hosting-coding-agents-on-a
- 用 Kiro CLI 自动搭建 FluentBit 日志采集方案:两种 EKS 埋点数据落地 S3 Parquet 的实战对比 | 亚马逊AWS官方博客 (2026-04-24)
- 将 Kiro CLI 封装为 REST API:双通道架构实践 | 亚马逊AWS官方博客 (2026-03-27)
- kiro-mcp-rds-mysql-upgrade
- 以Kiro快速部署云上Agent:只需几个小时,从业务需求到部署于Amazon Bedrock Agentcore落地 | 亚马逊AWS官方博客 (2026-04-02)
- litellm-amazon-quicksight-数据可视化配置手册
- litellm-生产级部署基于-aws-ecseks-的-ai-gateway-架构
- mcp-serveramazon-bedrock-agentcorequick-suite
- miro-amazon-bedrock-bug-routing
- monitor-and-debug-generative-ai-inference-with-sagemaker-det
- mountpoint-s3-与-s3-files-在-eks-上的实战对比
- Multi-Agent 架构在零售供应链运营中的实践:贯穿数据、洞察与行动 | 亚马逊AWS官方博客 (2026-04-14)
- Navigating EU AI Act Requirements for LLM Fine-Tuning
- netflix-real-time-service-topology
- network-firewall-deploy-guide-6-bedrock-ai-conflict-detection
- new-in-amazon-bedrock-agentcore-build-agents-with-broader-kn
- nvidia-nemotron-3-ultra-now-available-on-amazon-sagemaker-ju
- object-detection-with-amazon-nova-2-lite
- openai-models-and-codex-on-amazon-bedrock-are-now-generally-
- 把 OpenClaw 从个人助手变成客服:一次信任模型的翻转 | 亚马逊AWS官方博客 (2026-04-17)
- 给 Openclaw瘦身-利用Nova MME 和 S3 Vector实现Skill按需召回 | 亚马逊AWS官方博客 (2026-03-19)
- openclaw-multi-7-ecs-fargate-graviton
- OpenClaw 安全和功能增强实践 | 亚马逊AWS官方博客 (2026-03-12)
- 当 OpenClaw 学会”团队记忆”:一个面向多客户服务的企业级共享记忆系统设计 | 亚马逊AWS官方博客 (2026-04-17)
- optimize-blueprint-extraction-accuracy-in-amazon-bedrock-dat
- optimize-model-training-on-amazon-sagemaker-ai-with-nvidia-b
- 用 AI Agent 自动化日常办公工作流 — 在 AWS 上构建 Outlook 邮件助手的端到端实践
- powering-agentic-ai-sales-strategy-with-amazon-bedrock-agent
- 基于Strands Agents SDK和Amazon Bedrock AgentCore构建商品详情图广告词审查Agent | 亚马逊AWS官方博客 (2025-12-03)
- production-grade-ai-agents-for-financial-compliance-lessons-
- Prompting Amazon Nova 2 for content moderation
- 三剑合璧Quick Suite + Agent Core + Kiro联动实践:海外物流报价助手实战 | 亚马逊AWS官方博客 (2026-01-12)
- real-time-voice-agents-with-stream-vision-agents-and-amazon-nova-2-sonic
- reducing-container-cold-start-times-using-soci-index-on-dlam
- Restrict access to sensitive documents in your Amazon Quick knowledge bases for Amazon S3 | Amazon Web Services
- Restrict access to sensitive documents in your Amazon Quick knowledge bases for Amazon S3
- retrofit-dont-rebuild-agentic-overlays-for-transforming-lega
- running-comfyui-workflows-on-amazon-sagemaker-ai-processing
- 基于Strands框架和Bedrock AgentCore的SAP智能采购助手方案 | 亚马逊AWS官方博客 (2026-02-24)
- Scalable voice agent design with Amazon Nova Sonic: multi-agent, tools, and session segmentation
- scale-robot-reinforcement-learning-with-nvidia-isaac-lab-on-
- secure-ai-agents-with-policy-and-lambda-interceptors-in-amaz
- Securing AI agents: How AWS and Cisco AI Defense scale MCP and A2A deployments
- 基于AgentCore构建自学习、可进化的文旅行业近似信息抽取Agents | 亚马逊AWS官方博客 (2026-01-27)
- serverless-langgraph-multi-agent-aws
- stop-hand-tuning-kernels-how-neuron-agentic-development-acce
- 从0到1:联想基于Strands Agent SDK的资源智能巡检Agent创新 | 亚马逊AWS官方博客 (2026-02-24)
- strands-agents-cloud-cost-optimizer
- Build high-performance generative AI systems with Strands Agents
- 用 Strands Agents SDK 构建确定性数据分析:语义层 + VQR 在 Amazon Bedrock 上的实践 | 亚马逊AWS官方博客 (2026-04-16)
- the-art-and-science-of-hyperparameter-optimization-on-amazon-nova-forge
- transforming-rare-cancer-research-with-amazon-quick-integrat
- unlocking-ai-flexibility-in-europe-a-guide-to-cross-region-i
- 把 Kiro CLI 当作 Agent SDK:一键订阅即可构建你的Agent应用 | 亚马逊AWS官方博客 (2026-03-27)
- 使用 Kiro 规范驱动开发加速数据质量建设 | 亚马逊AWS官方博客 (2025-12-09)
- user-security-high-performance-bedrock-aws-china
- AI Agent 的迁移与现代化 — 使用 Amazon Bedrock AgentCore 将 OpenClaw 从单机改造为多租户 Serverless 架构 第一篇 | 亚马逊AWS官方博客 (2026-05-08)
- using-amazon-bedrock-agentcore-openclaw-multi-2
- AI Agent 的迁移与现代化 — 使用 Amazon Bedrock AgentCore 将 OpenClaw 从单机改造为多租户 Serverless 架构 第三篇 | 亚马逊AWS官方博客 (2026-05-08)
- AI Agent 的迁移与现代化 — 使用 Amazon Bedrock AgentCore 将 OpenClaw 从单机改造为多租户 Serverless 架构 第四篇 | 亚马逊AWS官方博客 (2026-05-08)
- using-amazon-bedrock-agentcore-openclaw-multi-5
- AI Agent 的迁移与现代化 — 使用 Amazon Bedrock AgentCore 将 OpenClaw 从单机改造为多租户 Serverless 架构 第六篇 | 亚马逊AWS官方博客 (2026-05-08)
- 使用Amazon EMR Serverless Storage简化运维节省成本 | 亚马逊AWS官方博客 (2025-12-16)
- 使用 Kiro CLI 和 Agent Client Protocol 构建飞书 AI 聊天机器人 | 亚马逊AWS官方博客 (2026-03-27)
- Valkey 为什么这么快?盘点 Valkey 中提升性能的黑科技
- verizon-connect-agentic-ai-100k-users
- 当 AI Agent 学会”忘记”:Amazon Bedrock AgentCore Memory 的记忆哲学” | 亚马逊AWS官方博客 (2026-03-04)
- Zenjoy 基于 Amazon Bedrock 和 EKS 构建 AIOps Agent:打通 Prometheus、ES 与夜莺的智能化告警实战
- 从 N 台 EC2 到 Amazon EKS + Amazon S3 Files:Waylens 的 OpenClaw 多智能体平台改造
- 使用-amazon-cloudfront-和-aws-waf-大规模交付-wordpress
- 使用-amazon-cognito-多区域复制提高应用程序韧性
- 使用-aws-network-firewall-服务审查-idc-和云上-vpc-间的流量-vgw-架构的设计和实验
- 使用-aws-security-agent-构建应用安全闭环从代码提交到漏洞修复的自动化之路
- 发布-amazon-bedrock-agentcore-web-搜索功能为人工智能代理提供实时准确的网络知识支撑
- 在 Amazon EC2 GPU 实例上部署 NVIDIA NemoClaw — 以 Amazon Bedrock 作为推理后端的生产级参考架构
- 在-amazon-eks-上使用-nvidia-gpu-operator-管理自定义-gpu-驱动与-cuda-工作负载
- 基于 Amazon Bedrock AgentCore 与 AWS DevOps Agent 打造对话式多账户运维助手
- 基于-amazon-ecs-fargate-自建-keycloak-作为-aws-iam-identity-center
- 基于-prowler-与-genai-构建金融行业智能合规中枢
- 存之有序治之有矩agent-记忆系统的工程实践与演进
- 宣布推出由-nvidia-rtx-pro-4500-blackwell-服务器版-gpu-加速的-amazon-ec2-
- 当-agentic-ai-重塑生产关系-智能体浪潮下的企业战略与行动框架
- 推出-amazon-bedrock-managed-knowledge-base助力企业人工智能应用程序更快速更准确
- 构建无服务器kiro调度平台用kiro-cli-eventbridge-ecs-fargate实现定时ai任务
- 现已推出由新的-aws-graviton5-处理器提供支持的-amazon-ec2-m9g-和-m9gd-实例
- 航班变更信息智能识别解决方案
- 规划-amazon-eks-从-132-升级到-135关键变更识别与逐版本实施路径
- 让-amazon-quick-操作飞书构建远程-mcp-服务的设计实践
- 试用-amazon-bedrock-中的新控制台体验该体验针对兼容-anthropic-和-openai-的-api-进
- 运行可全生命周期控制的隔离沙盒aws-lambda-推出-microvm
- 通过 LiteLLM 实现 Amazon Bedrock 成本管控:实时限额、多维监控与平台级兜底
- 飞来汇借助-aws-security-agent-构建跨境支付应用的智能安全防线
theregister.com(22 篇)¶
- Disgruntled researcher releases two more Microsoft zero-days
- Meta U-turns on encryption push for Instagram as DMs go plaintext
- 5237660-1
- Sovereign cloud is only possible if you’re Chinese or American: Gartner
- 5237875
- Anthropic’s bug-hunting Mythos was greatest marketing stunt ever, says cURL creator
- OpenAI buys AI consultancy to sell enterprises on its models
- anthropic-s-bug-hunting-mythos-was-greatest-marketing-stunt-ever-says-curl-creat
- disgruntled-researcher-microsoft-zero-days
- Japan’s PM orders cybersecurity review to defend against Anthropic Mythos
- London's police asked Big Tech for comms data over 700,000 times last year
- Malware crew TeamPCP open-sources its Shai-Hulud worm on GitHub
- Disgruntled researcher releases two more Microsoft zero-days
- Mozilla warns UK: Breaking VPNs will not magically fix Britain's age-check mess
- Mozilla warns UK: Breaking VPNs will not magically fix Britain's age-check mess
- Mozilla warns UK: Breaking VPNs will not magically fix Britain's age-check mess
- openai-buys-ai-consultancy-to-sell-enterprises-on-its-models
- vietnam-domestic-cloud
- vietnam-to-develop-domestic-cloud-so-it-can-ditch-risky-overseas-operators-for-g
- Vietnam to develop domestic cloud so it can ditch risky overseas operators for government workloads
- vietnam-to-develop-domestic-cloud
- Vietnam to develop domestic cloud
interconnects.ai(22 篇)¶
- claude-fable-5-and-new-ai-safety-fables
- finbarr-timbers-frontier-post-training-recipe-review-2026
- gemma-4-and-what-makes-an-open-model-succeed
- gpt-54-is-a-big-step-for-codex
- how-anthropic-vs-dow-impacts-open-interconnects
- How Open Model Ecosystems Compound
- Open and closed models are on different exponentials
- latest-open-artifacts-19-qwen-35-glm-5-minimax-25-chinese-la
- Latest open artifacts (#20): New orgs! New types of models! With Nemotron Super, Sarvam, Cohere Transcribe, & others
- latest-open-artifacts-21-open-model-bonanza-gemma-4-deepseek
- lossy-self-improvement
- my-bets-on-open-models-mid-2026
- nathan-lambert-claude-mythos-open-weights
- notes from inside chinas ai labs
- notes inside china ai labs lambert
- olmo-hybrid-and-future-llm-architectures
- reading-todays-open-closed-performance-gap
- Some ideas for what comes next, May 2026
- The distillation panic
- The inevitable need for an open model consortium
- What comes next with open models
- What I’ve been building: ATOM Report, post-training course, finishing my book, and ongoing research
oneusefulthing.org(21 篇)¶
- a-guide-to-which-ai-to-use-in-the-agentic-era
- against-brain-damage
- an-opinionated-guide-to-using-ai-right-now
- choosing-to-stay-human
- claude-code-and-what-comes-next
- claude-dispatch-and-the-power-of-interfaces
- co-existence-and-the-end-of-co-intelligence
- giving-your-ai-a-job-interview
- gpt-5-it-just-does-stuff
- management-as-ai-superpower
- mass-intelligence
- on-working-with-wizards
- What it feels like to work with Mythos (2026-06-09)
- real-ai-agents-and-real-work
- sign-of-the-future-gpt-55-mollick
- the-bitter-lesson-versus-the-garbage-can
- The recent history of AI in 32 otters
- the-shape-of-ai-jaggedness-bottlenecks-and-salients
- the-shape-of-the-thing
- three-years-from-gpt-3-to-gemini-3
- using-ai-right-now-a-quick-guide
huggingface.co(21 篇)¶
- Building Blocks for Foundation Model Training and Inference on AWS
- Unlocking asynchronicity in continuous batching
- Unlocking asynchronicity in continuous batching
- Build real agentic apps using CUGA: two dozen working examples on a lightweight harness
- CyberSecQwen-4B: Why Defensive Cyber Needs Small, Specialized, Locally-Runnable Models
- CyberSecQwen-4B
- Introducing the Ettin Reranker Family
- EVA-Bench Data 2.0: 3 Domains, 121 Tools, 213 Scenarios (2026-06-04)
- Fine-Tuning NVIDIA Cosmos Predict 2.5 with LoRA/DoRA for Robot Video Generation
- Fine-Tuning NVIDIA Cosmos Predict 2.5 with LoRA/DoRA for Robot Video Generation
- foundation-model-building-blocks
- Profiling in PyTorch (Part 2): From nn.Linear to a Fused MLP (2026-06-11)
- Introducing the Ettin Reranker Family
- molmomotion-language-guided-3d-motion-forecasting
- Nemotron 3.5 Content Safety: Customizable Multimodal Safety for Global Enterprise (2026-06-04)
- Fine-Tuning NVIDIA Cosmos Predict 2.5 with LoRA/DoRA for Robot Video Generation
- olmo-eval: An evaluation workbench for the model development
- servicenow vllm correctness huggingface
- Five labs, five minds: building a multi-model finance drama on small models (2026-06-09)
- vllm v0 to v1 correctness before corrections
- zai-org-GLM-5.2
mp.weixi(19 篇)¶
- Cursor 复盘 Harness:模型决定能力上限,Harness 决定生产下限
- Harness Engineering: 让 Coding Agent 可靠完成长程任务
- Harness Engineering:耗时一周,我是如何将应用的AI Coding率提升至90%的
- Harness不是目的,知识才是护城河 —— 一个AI工程交付团队的知识沉淀实践
- Karpathy 最新访谈:从 Vibe Coding 到 Agentic Engineering
- OpenClaw 完全指南:这可能是全网最新最全的系统化教程了!(3.2W字,建议收藏)
- RAG深度解析:分块、向量化、召回、重排,才是“蒸馏同事skill”的关键
- 一个文件让 AI Coding 效率翻倍:AGENTS.md 实践指南
- 一文带你弄懂 AI 圈爆火的新概念:Harness Engineering
- 两万字详解Claude Code源码核心机制
- 从Vibe Coding到Agentic Engineering:重构后台开发全流程
- 你不知道的 Agent:原理、架构与工程实践
- 天猫新品营销技术团队AI编码实战指南(上)
- 淘宝动效解决方案分享
- 深入理解 Claude Code 源码中的 Agent Harness 构建之道
- 深度拆解 Hermes Agent 记忆系统:它修正了 OpenClaw 的哪层误区?
- 看 AgentRun 如何玩转记忆存储,最佳实践来了!
- 范凌:当公司变成Agent,关于 AI 时代组织的 5 个反思
- 龙虾装上了,可以用来干啥?分享下我的 OpenClaw 多智能体团队搭建经验!
thehackernews.com(16 篇)¶
- 1-million-exposed-ai-services-hackernews
- ai-agents-inside-perimeter-hackernews
- cPanel, WHM Release Fixes for Three New Vulnerabilities — Patch Now
- Drupal to Release Urgent Core Security Updates on May 20, Sites Told to Prepare
- Funnel Builder Flaw Under Active Exploitation Enables WooCommerce Checkout Skimming
- Funnel Builder Flaw Under Active Exploitation Enables WooCommerce Checkout Skimming
- Funnel Builder Flaw Under Active Exploitation Enables WooCommerce Checkout Skimming
- GitHub Breached — Employee Device Hack Led to Exfiltration of 3,800+ Internal Repos
- Grafana GitHub Token Breach Led to Codebase Download and Extortion Attempt
- Grafana GitHub Token Breach Led to Codebase Download and Extortion Attempt
- Grafana GitHub Token Breach Led to Codebase Download and Ext
- interpol-operation-ramz-mena-cybercrime
- Microsoft Open-Sources RAMPART and Clarity to Secure AI Agents During Development
- GitHub Breached — Employee Device Hack Led to Exfiltration
- thehackernews-com-the-new-phishing-click-how-oauth
- Fake OpenAI Privacy Filter Repo Hits #1 on Hugging Face, Draws 244K Downloads
netflixtechblog.com(16 篇)¶
- a-human-augmenting-agentic-workflow-for-causal-inference
- democratizing-machine-learning-at-netflix-building-the-model
- dynamically-splitting-wide-partitions-in-cassandra-for-time-
- evaluating-netflix-show-synopses-with-llm-as-a-judge
- from-silos-to-service-topology-why-netflix-built-a-real-time
- high-throughput-graph-abstraction-at-netflix-part-i
- High-Throughput Graph Abstraction at Netflix: Part I
- how-netflix-simplified-batch-compute-with-kueue
- netflix-druid-interval-aware-caching
- netflix-live-operations-human-infrastructure
- scaling-archunit-with-nebula-archrules
- Scaling Camera File Processing at Netflix
- state-of-routing-in-model-serving
- thinking-fast-slow-for-a-personalized-notification-system
- toward-more-controllable-ai-video-editing-an-early-research-
- vmaf-v1-good-is-not-good-enough
arxiv.org(14 篇)¶
- A Bitter Lesson for Data Filtering
- agent skills teams architecture evolution selection guide
- Count Anything
- Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories
- arxiv-2606-12683-from-agi-to-asi
- arxiv-2606.17030
- Autodata: An agentic data scientist to create high quality synthetic data
- Mellum2 Technical Report
- Normalizing Trajectory Models
- Normalizing Trajectory Models
- qwen-image-flash-beyond-objective-design
- SkillOS: Learning Skill Curation for Self-Evolving Agents
- SkillOS
- SkillOS: Learning Skill Curation for Self-Evolving Agents
developer.nvidia.com(14 篇)¶
- building-ai-agents-for-ar-glasses-and-xr-devices-with-nvidia-xr-ai
- How to Automate AI Model Documentation with the NVIDIA MCG Toolkit
- nvidia agentic ai subsurface engineering
- Building for the Rising Complexity of Agentic Systems with Extreme Co-Design
- nvidia cut checkpoint costs nvcomp
- nvidia edge first llms av robotics
- nvidia extreme co design agentic systems
- nvidia gemma 4 edge ai
- nvidia gpu kernel translation cute python julia
- How to Automate AI Model Documentation with the NVIDIA MCG Toolkit
- nvidia multimodal rag knowledge systems
- nvidia nemotron 3 agents rag voice safety
- nvidia secure local agent nemoclaw openclaw
- nvidia telco reasoning models nemo
deepmind.google(11 篇)¶
- alphaevolve impact deepmind
- AlphaEvolve: Gemini-powered coding agent scaling impact across fields
- Reimagining the mouse pointer for the AI era
- deepmind-securing-future-ai-agents
- diffusiongemma-4x-faster-text-generation-deepmind-2026-06
- Introducing computer use in Gemini 3.5 Flash
- gemma-4-12b-deepmind-multimodal-2026-06
- We're launching the Google DeepMind Accelerator program in Asia Pacific to tackle environmental risks
- Introducing Gemini Omni
- investing-in-multi-agent-ai-safety-research
- powering-the-future-of-robotics-in-europe
anthropic.com(10 篇)¶
- 2028-two-scenarios-for-global-ai-leadership
- Anthropic Acquires Stainless
- anthropic-com-research-making-claude-a-chemist
- anthropic demystifying evals for ai agents
- anthropic multi agent research system
- anthropic natural language autoencoders
- claude-for-small-business
- claude-opus-47
- Introducing Claude for Small Business
- Introducing Claude Tag
testingcatalog.com(10 篇)¶
- Cline releases open-source agent runtime SDK
- Cline releases open-source agent runtime SDK
- Cline releases open-source agent runtime SDK
- Cline releases open-source agent runtime SDK
- Codex can now control other desktop devices via Computer Use
- Google’s Gemini Omni video model surfaces ahead of I/O debut
- Google’s Gemini Omni video model surfaces ahead of I/O debut
- Meta announced Muse Spark in Voice Mode and Meta Glasses
- Opensquilla Launches Open-Source AI Agent to Cut Token Costs
- opensquilla-launches-open-source-ai-agent-to-cut-token-costs
cio.com(9 篇)¶
- AI can write code, but the CIOs still owns the operating model
- AI-driven layoffs aren't making business sense
- AI-driven layoffs aren't making business sense
- CIOs rise to the global challenge
- Nearly every enterprise is investing in AI, but only 5% say their data is ready
- Nearly every enterprise is investing in AI, but only 5% say their data is ready
- www.cio-4171054-ai-driven-layoffs-arent-making-bus
- AI-driven layoffs aren't making business sense
- CEOs’ top priorities for IT leaders today
hackread.com(8 篇)¶
- AI Voice Cloning: The Technology Behind It, Who's Building It, and Where It's Headed
- Scammers Send Physical Phishing Letters to Steal Ledger Wallet Seed Phrases
- canvas-hackers-shinyhunters-say-their-official-domain-suspen
- canvas-hackers-shinyhunters-say-their-official-domain-was-suspended
- Fake Job Interview Apps Drop JobStealer Malware on Windows and macOS
- Romanian Man Faces Up to 30 Years in US Prison Over Vishing Scams
- Canvas Hackers ShinyHunters Say Their Official Domain Was Suspended
- TeamPCP Claims Sale of Mistral AI Repositories Amid Mini Shai-Hulud Attack
juejin.cn(8 篇)¶
- ai-skill-evolution底层逻辑
- ai-skill-skill-creator-源码拆解
- ai-skill-四层验证体系
- ai-skill-测评体系进阶指南
- ai-skill-测评报告解读
- ai-skill-测评指标体系
- ai-skill-测试用例设计
- harness-engineering-第三代工程范式
blog.crewai.com(7 篇)¶
- a-missing-layer-in-agentic-systems
- how-a-mid-tier-enterprise-saas-provider-automates-cloud-supp
- How we built Cognitive Memory for Agentic Systems
- lessons-from-2-billion-agentic-workflows
- stop-giving-your-agents-database-credentials
- your-first-ai-agent-should-do-one-thing-badly
- youre-building-agent-security-in-the-wrong-order
vercel.com(7 篇)¶
- AI Gateway production index
- automate-progressive-rollouts-with-vercel-flags-vercel
- Introducing deepsec: The security harness for finding vulnerabilities in your codebase
- Introducing eve
- Introducing Vercel Connect
- Protecting against token theft
- How Superset built the IDE for AI agents on Vercel
claude.com(7 篇)¶
- Building effective human-agent teams
- Claude Code 在大型代码库中的实战经验:从哪里入手?怎么做对?
- The evolution of agentic surfaces: building with Claude Managed Agents (Jun 10, 2026)
- claude managed agents official
- How Claude Code works in large codebases: Best practices and where to start
- How Claude Code works in large codebases: Best practices and where to start
- Running an AI-native engineering org
github.com(7 篇)¶
- browser harness github
- Graphsignal: LLM Inference Profiler
- Hunk - Review-first Terminal Diff Viewer
- Kipi: Open-source OSINT Investigation Platform with Autonomous Agent
- ml intern huggingface autonomous ml agent
- peerd: The first AI agent harness native to the browser
- pi mono github
ciodive.com(6 篇)¶
- 819775
- Engineering roles shift from developing code to managing AI
- OpenAI launches Daybreak to combat cyber threats
- Engineering roles shift from developing code to managing AI
- Engineering roles shift from developing code to managing AI
- U.S. Bank shifts critical apps to AWS for AI push | CIO Dive
microsoft.com(6 篇)¶
- Microsoft for Startups | Microsoft
- Defense at AI speed: Microsoft's new multi-model agentic security system
- New and improved Agent governance intelligent workflows
- Microsoft for Startups | Microsoft
- microsoft-startups
- Page not found | Microsoft Copilot Blog
infoworld.com(6 篇)¶
- Anthropic puts Claude agents on a meter across its subscriptions
- Anthropic puts Claude agents on a meter across its subscriptions
- The new AI lock-in
- new-lock
- The new AI lock-in
- Anthropic puts Claude agents on a meter across its subscriptions
a16z.news(6 篇)¶
- From “System of Record” to “System of Intelligence”
- From "System of Record" to "System of Intelligence"
- Is Software Losing Its Head?
- Is Software Losing Its Head?
- Is Software Losing Its Head?
- www-a16z-news-need-series-c-call-a16z
github.blog(6 篇)¶
- github agentic token efficiency
- github-blog-multilingual-ai-open-dataset
- Getting more from each token: How Copilot improves context handling and model routing
- GitHub Copilot Individual Plans: Flex Allotments in Pro and Pro+, and a New Max Plan
- github-secret-scanning-ai-ml-false-positive-reduction
- Improving token efficiency in GitHub Agentic Workflows
finovate.com(5 篇)¶
- 5 Things to Know about the CLARITY Act
- 5 Things to Know about the CLARITY Act
- clarity-act-5-things
- 5 Things to Know about the CLARITY Act
- Xero Announces Integration with Anthropic’s Claude
blog.cloudflare.com(5 篇)¶
- Bringing more agent harnesses and frameworks to Cloudflare, starting with Flue
- Announcing Claude Managed Agents on Cloudflare
- Project Glasswing: what Mythos showed us
- Unlocking the Cloudflare app ecosystem with OAuth for all
- Temporary Cloudflare Accounts for AI agents
lovable.dev(5 篇)¶
- Building is just the beginning: Introducing Discoverability
- Building is just the beginning: Introducing Discoverability | Lovable
- Building is just the beginning: Introducing Discoverability
- Building is just the beginning: Introducing Discoverability
- Turn repeated instructions into reusable skills in Lovable | Lovable
techzine.eu(5 篇)¶
- Checkmarx Jenkins plugin compromised in new supply chain attack
- servicenow-ui-is-dead-agent
- The UI is dead, long live the agent: ServiceNow goes headless and opens its platform
- the-ui-is-dead-long-live-the-agent-servicenow-goes-headless
- The UI is dead, long live the agent: ServiceNow goes headless and opens its platform
cursor.com(5 篇)¶
- Development environments for your cloud agents
- Development environments for your cloud agents
- Reward hacking is swamping model intelligence gains
- Introducing Composer 2.5
- Development environments for your cloud agents
cyberscoop.com(5 篇)¶
- google-amnesty-spyware-detection
- google-and-amnesty-international-teamed-up-to-make-android-s
- google-and-amnesty-international-teamed-up-to-make-it-harder
- A DOD contractor’s API flaw exposed military course data and service member records
- White House cyber official: identity security matters more
tomtunguz.com(5 篇)¶
- Localmaxxing
- Intelligence Per Dollar
- The Unsustainable Subsidy
- tomtunguz-com-so-you-want-to-sell-inference
- Software After AI
depthfirst.com(5 篇)¶
- NGINX Rift: Achieving NGINX Remote Code Execution via an 18-Year-Old Vulnerability | depthfirst
- NGINX Rift: Achieving NGINX Remote Code Execution via an 18-Year-Old Vulnerability
- Open Defense Initiative | depthfirst
- Open Defense Initiative
- Open Defense Initiative | depthfirst
stochasticparrot.substack.com(5 篇)¶
- stochastic parrot deep mystery llms
- stochastic parrot language models and meaning
- stochastic parrot marcus ai productivity
- stochastic parrot thought experiment
- while breathless in stodgy viridian
networkworld.com(5 篇)¶
- Versa takes aim at fragmented enterprise security with CSPM, orchestration update, and AI agent controls
- wi-fi-8-closer-than-you-think
- wi-fi-8-is-closer-than-you-think
- Versa takes aim at fragmented enterprise security with CSPM, orchestration update, and AI agent controls
- www-networkworld-com-wi-fi-8-is-closer-than-you-think
venturebeat.com(4 篇)¶
- AI tool poisoning exposes a major flaw in enterprise agent security
- Claude's next enterprise battle is not models: it's the agent control plane
- Intercom, now called Fin, launches an AI agent whose only job is managing another AI agent
- Perceptron Mk1 shocks with highly performant video analysis AI model 80-90% cheaper than Anthropic, OpenAI & Google
thewhiteh4t.github.io(4 篇)¶
- LLMReaper - DOM Based AI Conversation Exfiltration via Browser Extensions
- LLMReaper - DOM Based AI Conversation Exfiltration via Browser Extensions
- LLMReaper - DOM Based AI Conversation Exfiltration via Browser Extensions
- LLMReaper - DOM Based AI Conversation Exfiltration via Browser Extensions
cnbc.com(4 篇)¶
- Americans still feel pessimistic about the economy. When will it change?
- Consumer Confidence
- Nvidia embraces role of AI investor, pushing past $40 billion in equity bets this year
- nvidias-jensen-huang-bets-on-this-british-startup-to-build-next-frontier-of-ai
jiaxin-wen.github.io(4 篇)¶
- Generalization Dynamics of LM Pre-training — Jiaxin Wen
- Generalization Dynamics of LM Pre-training — Jiaxin Wen
- Generalization Dynamics of LM Pre-training — Jiaxin Wen
- generalization-dynamics-pre-training-jiaxin-wen
llm-as-a-verifier.notion.site(4 篇)¶
- LLM-as-a-Verifier: A General-Purpose Verification Framework
- llm-as-a-verifier-a-general-purpose-verification
- LLM-as-a-Verifier: A General-Purpose Verification Framework
- LLM-as-a-Verifier: A General-Purpose Verification Framework
pytorch.org(4 篇)¶
- pytorch-2-12-release-blog
- PyTorch 2.12 Release Blog
- pytorch in kernel recsys optimization
- PyTorch 2.12 Release Blog – PyTorch
technologyreview.com(3 篇)¶
- Cyberscammers are bypassing banks’ security with illicit tools sold on Telegram
- Cyberscammers are bypassing banks' security with illicit tools sold on Telegram
- Cyberscammers are bypassing banks’ security with illicit tools sold on Telegram
linear.app(3 篇)¶
back.engineering(3 篇)¶
- Static Devirtualization of Themida
- Static Devirtualization of Themida
- Static Devirtualization of Themida
fandf.co(3 篇)¶
- Archera • Insured cloud commitments for AWS, Azure, and Google
- identity-behavior-context-itdr-solution-teleport
- identity-behavior-context-itdr-solution
figma.com(3 篇)¶
- 4 ways we’re using our MCP server at Figma
- What the design-to-code loop unlocks
- Figma Make, now on your local code
computerworld.com(3 篇)¶
- 5 ways to curb AI sprawl without stifling innovation
- Google brings local AI agents to laptops with Gemma 4 12B
- Microsoft to retire ‘Together Mode,’ its virtual meeting space for Teams
pymnts.com(3 篇)¶
- Amazon Turns Alexa Into Its Next Storefront
- Thrive Capital Bets $100 Million on Shopify’s AI Future
- Chime Turns a Profit as Members Hit 10.2 Million
projectzero.google(3 篇)¶
- A 0-click exploit chain for the Pixel 10: When a Door Closes, a Window Opens
- A 0-click exploit chain for the Pixel 10: When a Door Closes, a Window Opens
- A 0-click exploit chain for the Pixel 10: When a Door Closes, a Window Opens
theglobeandmail.com(3 篇)¶
- Affirm Maps Road to $100B GMV With Card, AI Commerce and Global Expansion
- affirm-maps-road-to-100b-gmv-with-card-ai-commerce
- Affirm Maps Road to $100B GMV With Card, AI Commerce and Global Expansion
cloud.google.com(3 篇)¶
- Agent Executor, Google’s distributed Agent Runtime
- Agent Executor, Google’s distributed Agent Runtime
- Adversaries Leverage AI for Vulnerability Exploitation, Augmented Operations, and Initial Access
blog.kilo.ai(3 篇)¶
- agentic-engineering-leadership
- deepseek-v4-pro-vs-claude
- We Tested DeepSeek V4 Pro and Flash Against Claude Opus 4.7
wiz.io(3 篇)¶
- A Framework for AI Threat Readiness
- Fragnesia: Linux Kernel Local Privilege Escalation via ESP-in-TCP
- www-wiz-io-mini-shai-hulud-strikes-again-tanstack-more-npm-packages-compromised
developers.googleblog.com(3 篇)¶
- Announcing Genkit Middleware: Intercept, extend, and harden your agentic apps
- announcing-genkit-middleware-intercept-extend-and-harden-your-agentic-apps
- Announcing Genkit Middleware
paymentsdive.com(3 篇)¶
- block-leans-into-its-ai-future-payments-dive
- senators-bnpl-credit-bureaus
- senators-query-credit-bureaus-on-bnpl-1
blogs.nvidia.com(3 篇)¶
- Vera Arrives: NVIDIA’s First CPU Built for Agents Lands at Top AI Labs
- nvidia-blackwell-mlperf-training-6-0-benchmark-results-2026-06
- Vera Arrives: NVIDIA’s First CPU Built for Agents Lands at Top AI Labs
prismml.com(3 篇)¶
- Introducing 1-bit and Ternary Bonsai Image 4B: Image Generation for Local Devices
- Introducing 1-bit and Ternary Bonsai Image 4B: Image Generation for Local Devices
- Introducing 1-bit and Ternary Bonsai Image 4B: Image Generation for Local Devices
creativeboom.com(3 篇)¶
- Have your views on AI changed
- Linn Fritz looks at the lighter side of life
- this-redesign-of-stokes-coffee-is-a-masterclass-in-change-everything-but-don-t-c
seangoedecke.com(3 篇)¶
- DeepSeek-V4-Flash means LLM steering is interesting again
- AI GPUs probably live longer than three years
- Build agents, not pipelines
blog.google(3 篇)¶
- DiffusionGemma: 4x faster text generation (2026-06-10T16:00:00+00:00)
- gemma 4 multi token prediction drafters
- Gemma 4 QAT Models: Optimizing model compression for mobile and laptop efficiency (2026-06-05)
americanbanker.com(3 篇)¶
- Fed Data Shows Cash Use Holding Steady
- late-student-loan-payments-are-rising-is-a-domino-effect-com
- late-student-loan-payments-are-rising-is-a-domino-effect-coming
letsdatascience.com(3 篇)¶
- Google Debuts Gemini-Focused Updates at I/O 2026
- Igor Babuschkin Seeks Up To $1 Billion For River AI
- Igor Babuschkin Seeks Up To $1 Billion For River AI
inngest.com(3 篇)¶
- Inngest - AI and backend workflows, orchestrated at any scale
- Inngest - AI in Production: The 2026 Benchmark Report
- Inngest - AI in Production: The 2026 Benchmark Report
news.sap.com(3 篇)¶
- news-sap-com-sap-unveils-the-autonomous-enterprise
- sap-unveils-autonomous-enterprise
- SAP Unveils the Autonomous Enterprise
heyfuturenexus.com(3 篇)¶
- Private Fintech Has Quietly Become Bigger Than Public Fintech. Now What?
- Private Fintech Has Quietly Become Bigger Than Public Fintech. Now What?
- Private Fintech Has Quietly Become Bigger Than Public Fintech. Now What?
securityaffairs.com(3 篇)¶
- Pwn2Own Berlin 2026, Day Three: DEVCORE Crowned Master of Pwn, $1.298 Million Total
- Hackers accessed BWH Hotels reservation system for months
- ShinyHunters hack 7-Eleven: franchisee data and Salesforce records exposed
saastr.com(3 篇)¶
- "20VC x SaaStr: The Most Aggressive Quarter in American Capitalism, Palantir’s Ru"20VC x SaaStr: The Most Aggressive Quarter in American Capitalism, Palantir’s Ru
- Who Winning Enterprise AI Now
- When Growth Slows, Is It Sales' Fault or the Product's Fault? The Answer Has Changed.
cryptobriefing.com(3 篇)¶
- tether-developer-grants-local-ai
- tether-launches-developer-grants-program-for-local-ai-paymen
- Tether launches developer grants program for local-first AI and payments infrastructure
volcengine.com(3 篇)¶
zeroheight.com(2 篇)¶
- 10 Common Component Architecture Mistakes in Figma Design Systems
- 10 common component architecture mistakes in Figma design systems
threadreaderapp.com(2 篇)¶
cncf.io(2 篇)¶
- Building Jaeger’s ClickHouse backend: 8.6× compression on 10 million spans
- Automating Confidential Containers (CoCo) infrastructure with Kyverno
cybersecuritydive.com(2 篇)¶
- New cybersecurity industry coalition aims to lead US critical infrastructure protection
- CISA urges critical infrastructure firms to ‘fortify’ before it’s too late | Cybersecurity Dive
helloworldmedia.notion.site(2 篇)¶
dxlong2000.github.io(2 篇)¶
- a2rd-agentic-autoregressive-diffusion-long-video
- A²RD: Agentic Autoregressive Diffusion for Long Video Consistency
pages.awscloud.com(2 篇)¶
qgallouedec-tito.hf.space(2 篇)¶
maybern.com(2 篇)¶
- Why Internally-Built AI Fails Fund Accounting Audits
- Why Internally-Built AI Fails Fund Accounting Audits
airbyte.com(2 篇)¶
alphaxiv.org(2 篇)¶
red.anthropic.com(2 篇)¶
- Anthropic LLM ATT&CK Navigator: AI-Enabled Cyber Operations
- Anthropic N-days: Frontier Agent Vulnerability Research
theguardian.com(2 篇)¶
- Anthropic to share Mythos cyber flaw findings with global finance watchdog
- Anthropic to share Mythos cyber flaw findings with global finance watchdog
skids.dev(2 篇)¶
- Tokenomics: the 62.5-minute rule for Claude's cache
- Tokenomics: the 62.5-minute rule for Claude's cache
williamangel.net(2 篇)¶
blog.zsec.uk(2 篇)¶
cyera.com(2 篇)¶
- bleeding-llama-critical-unauthenticated-memory-leak-in-ollama
- Claw Chain: Cyera Research Unveil Four Chainable Vulnerabilities in OpenClaw
developers.openai.com(2 篇)¶
- build-live-translation-apps-with-gpt-realtime-translate
- skills anthropic openai comparison frontend design
onlycfo.io(2 篇)¶
krebsonsecurity.com(2 篇)¶
searchengineland.com(2 篇)¶
- Inside ChatGPT Search: how web.run and fan-out queries shape results
- Google adds llms.txt check to Chrome Lighthouse
citriniresearch.com(2 篇)¶
x.com(2 篇)¶
- claude code core developer lessons action space design
- [claude-code开发负责人-为何放弃rag而选择agentic-search](https://x.com/bcherny/status/2017824286489383315
作为一个被)
thezvi.wordpress.com(2 篇)¶
hs.playerzero.ai(2 篇)¶
- Code Simulation for Enterprise Engineering | PlayerZero
- Code Simulation for Enterprise Engineering | PlayerZero
cside.com(2 篇)¶
ipurple.team(2 篇)¶
exaforce.ai(2 篇)¶
weandthecolor.com(2 篇)¶
- Freelance Designers Can't Compete With a $20/Month AI Subscription - Here's What Actually Works Now
- Top 10 Design Gadgets for Creative Professionals That Will Change How You Work in 2026
gbhackers.com(2 篇)¶
- Sandworm Hackers Shift From IT Breaches to Critical OT Targets
- Sandworm Hackers Shift From IT Breaches to Critical OT Targets
workspaceupdates.googleblog.com(2 篇)¶
- google-workspace-updates-small-businesses-can-now-import-use
- workspaceupdates-googleblog-com-google-workspace-updates
memory.cobanov.dev(2 篇)¶
semgrep.dev(2 篇)¶
stratechery.com(2 篇)¶
huntress.com(2 篇)¶
- How Unified EDR and ITDR Stop Attacks Before They Spread
- How Unified EDR and ITDR Stop Attacks Before They Spread
implicator.ai(2 篇)¶
thinkingmachines.ai(2 篇)¶
bishopfox.com(2 篇)¶
manus.im(2 篇)¶
docs.smith.langchain.com(2 篇)¶
latent.space(2 篇)¶
lightfield.app(2 篇)¶
lightseek.org(2 篇)¶
kaspersky.com(2 篇)¶
salesforce.com(2 篇)¶
- Meet Customers Where They Are: Agentforce Contact Center Now Offers WhatsApp Voice
- CEOs of the classroom: Why principals are the key to the AI era
runtimewire.com(2 篇)¶
originhq.com(2 篇)¶
- MXC Internals: How Microsoft's eXecution Containers Actually Isolate Agent Code
- Mxc Execution Containers Internals Origin
xbow.com(2 篇)¶
tanstack.com(2 篇)¶
- Postmortem: TanStack npm supply-chain compromise | TanStack Blog
- postmortem-tanstack-npm-supply-chain-compromise-tanstack-blog
tanayj.com(2 篇)¶
productledseo.com(2 篇)¶
research.perplexity.ai(2 篇)¶
workos.com(2 篇)¶
programbench.com(2 篇)¶
magazine.sebastianraschka.com(2 篇)¶
- Recent Developments in LLM Architectures: KV Sharing, mHC, and Compressed Attention
- Recent Developments in LLM Architectures: KV Sharing, mHC, a
status.rubygems.org(2 篇)¶
labs.reversec.com(2 篇)¶
- Skill Issues: Compromising Claude Code with malicious skills & agents -- Part 1
- Skill Issues: Compromising Claude Code with malicious skills & agents — Part 1
stackoverflow.blog(2 篇)¶
- Announcing Stack Overflow for Agents (2026-06-10)
- The New Bottleneck: Theory of Constraints in the Age of AI Coding
temporal.io(2 篇)¶
fintechbrainfood.com(2 篇)¶
- The stablecoin 24x7 money loop: business case for banks
- 🧠 The Token Economy pt2: The Intelligence Company Gets Built
datadoghq.com(2 篇)¶
- Toto 2: Context-aware log analytics for complex distributed systems
- Toto 2.0: Time series forecasting enters the scaling era
unit42.paloaltonetworks.com(2 篇)¶
- Tracking TamperedChef Clusters via Certificate and Code Reuse
- Tracking TamperedChef Clusters via Certificate and Code Reuse
clearseclabs.com(2 篇)¶
- weve-been-here-before-ai-vulnerability-research
- We've Been Here Before: Decompilers, Fuzzers, and Now AI
asymco.com(1 篇)¶
schneier.com(1 篇)¶
quarter--mile.com(1 篇)¶
1password.com(1 篇)¶
notes.jim-nielsen.com(1 篇)¶
maeda.pm(1 篇)¶
jamesoclaire.com(1 篇)¶
wasp.sh(1 篇)¶
9to5google.com(1 篇)¶
bitwarden.com(1 篇)¶
calcalistech.com(1 篇)¶
elenajs.com(1 篇)¶
laurent.le-brun.eu(1 篇)¶
gowers.wordpress.com(1 篇)¶
a16z.com(1 篇)¶
ixdf.org(1 篇)¶
adobe.design(1 篇)¶
trevorlasn.com(1 篇)¶
afine.com(1 篇)¶
samihonkonen.com(1 篇)¶
antfly.io(1 篇)¶
picrew.github.io(1 篇)¶
addyosmani.com(1 篇)¶
learn.thedesignsystem.guide(1 篇)¶
bcs.org(1 篇)¶
snowflake.com(1 篇)¶
softwaredoug.com(1 篇)¶
tremendous.blog(1 篇)¶
lennysnewsletter.com(1 篇)¶
mitiga.io(1 篇)¶
oracle.com(1 篇)¶
noma.security(1 篇)¶
cloudbees.com(1 篇)¶
blog.talosintelligence.com(1 篇)¶
terriblesoftware.org(1 篇)¶
humansecurity.com(1 篇)¶
d2908q01vomqb2.cloudfront.net(1 篇)¶
almcorp.com(1 篇)¶
aboutamazon.com(1 篇)¶
platform.claude.com(1 篇)¶
blog.hello.coop(1 篇)¶
security.apple.com(1 篇)¶
algolia.com(1 篇)¶
arcticwolf.com(1 篇)¶
thelettertwo.com(1 篇)¶
blackhillsinfosec.com(1 篇)¶
automationanywhere.com(1 篇)¶
app.zuddl.com(1 篇)¶
app.sciencesays.com(1 篇)¶
tierfs.com(1 篇)¶
grafana.com(1 篇)¶
blog.himanshuanand.com(1 篇)¶
vlt.io(1 篇)¶
blog.tubikstudio.com(1 篇)¶
varonis.com(1 篇)¶
brethorsting.com(1 篇)¶
browser-use.com(1 篇)¶
blog.crossplane.io(1 篇)¶
martinfowler.com(1 篇)¶
blog.calif.io(1 篇)¶
calv.info(1 篇)¶
techcommunity.microsoft.com(1 篇)¶
cilium.io(1 篇)¶
- Kubernetes Runtime Security: The Guide (2026-06-07)
nextplatform.com(1 篇)¶
brainoverflow.blog(1 篇)¶
- Hidden Gaps in Claude Code Security Reviews (2026-06-01T11:35:14-07:00)
cohere.com(1 篇)¶
unchainedcrypto.com(1 篇)¶
reflex.dev(1 篇)¶
computerweekly.com(1 篇)¶
doubao.com(1 篇)¶
docker.com(1 篇)¶
cointelegraph.com(1 篇)¶
cybersecurityreach.org(1 篇)¶
darioamodei.com(1 篇)¶
- Dario Amodei — Policy on the AI Exponential (2026-06-12)
techmonitor.ai(1 篇)¶
databricks.com(1 篇)¶
- Announcing the Databricks storage ecosystem: Governing the enterprise data estate, wherever it lives
securitylabs.datadoghq.com(1 篇)¶
writing.nikunjk.com(1 篇)¶
dev.jimgrey.net(1 篇)¶
unknownarts.co(1 篇)¶
tympanus.net(1 篇)¶
uxmag.com(1 篇)¶
research.eye.security(1 篇)¶
inferterra.com(1 篇)¶
digitalocean.com(1 篇)¶
discord.com(1 篇)¶
shubs.io(1 篇)¶
heyitsas.im(1 篇)¶
dropbox.tech(1 篇)¶
nesbitt.io(1 篇)¶
eclecticlight.co(1 篇)¶
allenai.org(1 篇)¶
idlemachines.co.uk(1 篇)¶
thestateofbrand.com(1 篇)¶
refactoringenglish.com(1 篇)¶
usefastlane.ai(1 篇)¶
helpnetsecurity.com(1 篇)¶
thefinanser.com(1 篇)¶
siliconangle.com(1 篇)¶
boagworld.com(1 篇)¶
georg-schwarz.com(1 篇)¶
slack.engineering(1 篇)¶
cognition.ai(1 篇)¶
fusedash.ai(1 篇)¶
simonwillison.net(1 篇)¶
credrelay.com(1 篇)¶
ghostbyt3.github.io(1 篇)¶
businessinsider.com(1 篇)¶
socket.dev(1 篇)¶
unsloth.ai(1 篇)¶
seancai.com(1 篇)¶
goodfire.ai(1 篇)¶
research.google(1 篇)¶
docs.cloud.google.com(1 篇)¶
gptomics.com(1 篇)¶
whitefiber.com(1 篇)¶
go.granola.ai(1 篇)¶
optro.ai(1 篇)¶
greptile.com(1 篇)¶
nathan.rs(1 篇)¶
thisiscolossal.com(1 篇)¶
hacktivis.me(1 篇)¶
heidihealth.com(1 篇)¶
worksonmymachine.ai(1 篇)¶
zhuanlan.zhihu.com(1 篇)¶
leehanchung.github.io(1 篇)¶
itamargilad.com(1 篇)¶
docusign.com(1 篇)¶
lesswrong.com(1 篇)¶
blog.bytebytego.com(1 篇)¶
vincenzoiozzo.com(1 篇)¶
monokai.com(1 篇)¶
0xkato.xyz(1 篇)¶
- How LLMs Actually Work (2026-06-01)
read.technically.dev(1 篇)¶
newsletter.eng-leadership.com(1 篇)¶
thesaascfo.com(1 篇)¶
webflow.com(1 篇)¶
questdb.com(1 篇)¶
sparktoro.com(1 篇)¶
posthog.com(1 篇)¶
quantamagazine.org(1 篇)¶
pulse2.com(1 篇)¶
hongkiat.com(1 篇)¶
image-to-video.ai(1 篇)¶
incendium.rocks(1 篇)¶
tech.instacart.com(1 篇)¶
blog.mozilla.ai(1 篇)¶
developer.mozilla.org(1 篇)¶
blog.sentry.io(1 篇)¶
nautil.us(1 篇)¶
blog.janestreet.com(1 篇)¶
blog.jetbrains.com(1 篇)¶
npr.org(1 篇)¶
k-dense.ai(1 篇)¶
kasra.blog(1 篇)¶
klarna.com(1 篇)¶
kristoff.it(1 篇)¶
langchain.com(1 篇)¶
lesecretairedefernand.co(1 篇)¶
developer.chrome.com(1 篇)¶
nousresearch.com(1 篇)¶
liquid.ai(1 篇)¶
alignmentforum.org(1 篇)¶
parsiya.net(1 篇)¶
pushsecurity.com(1 篇)¶
lmsys.org(1 篇)¶
developer.okta.com(1 篇)¶
lucasfcosta.com(1 篇)¶
a16z.simplecast.com(1 篇)¶
markupextension.com(1 篇)¶
blog.modelcontextprotocol.io(1 篇)¶
corememory.com(1 篇)¶
thenextweb.com(1 篇)¶
mistral.ai(1 篇)¶
modal.com(1 篇)¶
aifoc.us(1 篇)¶
welodata.ai(1 篇)¶
daniel.haxx.se(1 篇)¶
n8n.io(1 篇)¶
injuly.in(1 篇)¶
theverge.com(1 篇)¶
straiker.ai(1 篇)¶
dwarkesh.com(1 篇)¶
techcrunch.com(1 篇)¶
notnotp.com(1 篇)¶
novee.security(1 篇)¶
research.nvidia.com(1 篇)¶
byungkwanlee.github.io(1 篇)¶
brandonrozek.com(1 篇)¶
alignment.openai.com(1 篇)¶
openai.com(1 篇)¶
warp.dev(1 篇)¶
news.aakashg.com(1 篇)¶
elenaverna.com(1 篇)¶
wersm.com(1 篇)¶
plaid.com(1 篇)¶
platformer.news(1 篇)¶
playerzero.ai(1 篇)¶
lilianweng.github.io(1 篇)¶
blog.grandimam.com(1 篇)¶
linuxfoundation.org(1 篇)¶
webdirections.org(1 篇)¶
mrmarket.bearblog.dev(1 篇)¶
rajveerbachkaniwala.com(1 篇)¶
randsinrepose.com(1 篇)¶
growth-memo.com(1 篇)¶
recursive.com(1 篇)¶
recyclebin.zip(1 篇)¶
ffnews.com(1 篇)¶
welivesecurity.com(1 篇)¶
role-confusion.github.io(1 篇)¶
roman.pt(1 篇)¶
blog.samanl33t.com(1 篇)¶
pathtostaff.com(1 篇)¶
try.digitalocean.com(1 篇)¶
sentinelone.com(1 篇)¶
slcyber.io(1 篇)¶
blog.jim-nielsen.com(1 篇)¶
engineering.atspotify.com(1 篇)¶
brandsynario.com(1 篇)¶
boringsql.com(1 篇)¶
squareup.com(1 篇)¶
stability.ai(1 篇)¶
stripe.com(1 篇)¶
sunilpai.dev(1 篇)¶
sysdig.com(1 篇)¶
clearbluejar.github.io(1 篇)¶
tantosec.com(1 篇)¶
newatlas.com(1 篇)¶
tenorshare.ai(1 篇)¶
hyperdimensional.co(1 篇)¶
fin.ai(1 篇)¶
drata.com(1 篇)¶
ericwbailey.website(1 篇)¶
dead.online(1 篇)¶
lucumr.pocoo.org(1 篇)¶
tines.com(1 篇)¶
danielmiessler.com(1 篇)¶
tidemarkcap.com(1 篇)¶
theburningmonk.com(1 篇)¶
tonsky.me(1 篇)¶
blog.trailofbits.com(1 篇)¶
- The sorry state of skill distribution (2026-06-03)
utoronto.ca(1 篇)¶
nrehiew.github.io(1 篇)¶
dylanzsz.github.io(1 篇)¶
cms.greatquestion.co(1 篇)¶
uxcrush.com(1 篇)¶
varoa.net(1 篇)¶
ref.getviktor.com(1 篇)¶
blog.ammaraskar.com(1 篇)¶
andonlabs.com(1 篇)¶
webkit.org(1 篇)¶
thisisgoingtobebig.com(1 篇)¶
blog.murphytrueman.com(1 篇)¶
tomassetti.me(1 篇)¶
designbeep.com(1 篇)¶
martech.org(1 篇)¶
blog.yaelwrites.com(1 篇)¶
vas-blog.pages.dev(1 篇)¶
blogs.vmware.com(1 篇)¶
vinibrasil.com(1 篇)¶
openclaw.ai(1 篇)¶
informationweek.com(1 篇)¶
jorijn.com(1 篇)¶
labs.watchtowr.com(1 篇)¶
semiengineering.com(1 篇)¶
- Wi-Fi Flies Higher As Edge AI Build-Out Takes Root (2026-06-15)
referenceboard.app(1 篇)¶
sans.org(1 篇)¶
bettercloud.com(1 篇)¶
blocksandfiles.com(1 篇)¶
yoonholee.com(1 篇)¶
- We Should Take Text Optimization More Seriously (2026-06-08)