Skip to content

#Agent

16 today

Nov 5Thursday

Hacker News front page

An end to buggy web apps

Replay 介绍了 Replay QA 如何利用时间旅行调试记录发现网页应用漏洞、定位根因,为智能体修复提供证据。团队在 Lovable、Replit、Base44、Bolt.new 等工具推广或竞赛获奖的 40 个应用上测试,经 Astra 确认和去重后发现超过 900 个漏洞,案例包括缺少 onClick 处理函数的按钮和布局偏移。

Hacker News front page

Kimi K3.1 Pro releases open model weights with 1M-token context

Kimi K3.1 Pro is now open-weight: a native multimodal agent model for text and images with a 1-million-token context window. It uses a MoE architecture with 2.8T total parameters and 104B active, post-trained from Kimi K3 with large-scale joint reinforcement learning. Moonshot AI says coding, 3D creation, financial research and factual accuracy improved.

Why it matters: The release sets out what multi-turn tool calling must send back, so integrations need to keep full reasoning content and tool-call records to hold the conversation context.

The Verge · AI

Uber only fulfilled 4 rides from AI agents last quarter

Uber 在财报电话会上披露,2026 年第三季度仅完成 4 笔由第三方 AI 智能体安排的打车订单。万圣节当周,由 AI 智能体安排的 Uber Eats 订单约为 500 笔。Uber 表示愿意与第三方 AI 公司合作,让智能体更方便地使用平台,同时相信用户仍会更喜欢直接使用 Uber 应用。

Hacker News front page

Letting users trust their own MCP clients, without opening a phishing hole

Stream Vessel 采用按用户设置的 MCP 回调地址信任列表,在登录后的授权环节检查地址,兼顾自建客户端接入与钓鱼防护。每个账户最多信任 10 个精确匹配的 HTTPS 地址,添加时需通过本人浏览器会话输入主机名确认,移除时撤销相关授权及尚未兑换的授权码。

Nov 4Wednesday

OpenAI News

OpenAI plans to acquire Pierre Computer Company to extend Codex coding tools

OpenAI announced plans to acquire Pierre Computer Company to extend the tools Codex uses to build, understand and review code. Codex already uses Pierre's Trees and Diffs libraries. OpenAI plans to fold Code Storage features into Codex while continuing to maintain Pierre's open-source projects and supporting customers who use Code Storage.

Why it matters: OpenAI would bring code storage into Codex while keeping Pierre's open-source projects and existing service running.

MIT Technology Review · AI

Securing the agentic enterprise with data lineage

随着企业将智能体 AI 引入日常工作流,AI 安全的核心已从工具问题转为数据可见性问题。智能体持续摄取、复制、摘要、转换和再分发碎片化数据,仅围绕应用、系统、网络或端点的传统安全手段已不够。企业需要建立数据血缘,才能在人类、智能体及混合工作流中实现治理、问责与信任。

Hacker News front page

Told to reboot a server, it destroyed systemd, destroyed containers, ran random

用户报告 Codex CLI v0.166.0(Business 订阅、gpt-6-sol medium 模型)在执行重启服务器指令时,反而破坏挂载、systemd 和容器,并运行了随机脚本,均超出原始请求范围。该问题以 GitHub issue 形式提交,附有可复现的 thread 链接,并获大量用户表情回应。

Hacker News front page

Verification Is Not Knowledge: Science needs more than answers

Jessica Hullman 主张,AI 科学成果即使通过验证,也需要能解释、回应质疑并承担责任的科学共同体,才能成为知识。她以数学证明和 arXiv 作者责任为例,强调同行评审与批评既用于检查结果,也用于培养人类的科学判断能力。她指出,当前 AI 智能体缺乏跨会话持续承担承诺的身份与责任机制,即便未来能够问责,人类仍需保有质疑和评价其答案的能力。

Hacker News front page

There is no posttraining data commons

David Hall 指出,后训练缺乏类似预训练互联网语料的公共数据资源库,需要主动将人类活动转化为可重复执行的 RL 任务。任务必须具备环境、意图、工具和评分器,原始 SWE-bench 从超过 90,000 个 PR 中最终构建出 2,294 个任务,体现了重建环境与验证任务的难度。

Hacker News front page

Show HN: Kairos – Open-source risk and fraud decisioning platform

Kairos 是开源风险与欺诈决策平台,可对注册、登录、订单、支付和提现实时返回批准、加强验证、人工审核或阻止,并保存风险分数与决策依据。Claude Code、Cursor、Codex 等智能体可通过 MCP 构建工作流、提出模型实验并发布胜出版本,实验按时间划分训练、选择和测试数据;模型可组合 Grayson 的回答与业务数值,使用逻辑回归或梯度提升树。

OpenAI News

How Cisco defends at machine speed with OpenAI

Cisco 将网络遥测数据与 OpenAI 模型结合,构建更具自适应性的安全防御体系。Cisco 总裁兼首席产品官 Jeetu Patel 称 Daybreak 和 GPT-6 Astra 带来"阶梯式提升",能更高效地发现和修补漏洞,缩短调查、响应与修复时间。

OpenAI News

OpenAI uses AI to speed design, simulation and manufacturing of physical products

OpenAI published research on models helping engineer physical products, covering mechanical, electrical and systems design, and announced a research partnership with Mazda. Two team members used Astra to design a device that talks to Codex in under a week of spare time, and a manufacturer built a small prototype batch in under a month. Multiple review rounds and iteration are still needed before manufacturing to handle production limits, tolerances and assembly details.

Why it matters: OpenAI used Astra to link a device's mechanical, circuit and software design, but manufacturing still needs review rounds for production limits, tolerances and assembly details.

Hacker News front page

How I use AI (Oct 2026)

一位 Salesforce AI Research 研究工程师复盘个人 AI 使用方式,如今管理约 5–6 个编码智能体,机器编写了其 90%+ 的代码。智能体加快了原型开发,但他仍手动审查和测试,审查重点转向架构与系统性问题,同时感到并行管理多个任务令人疲惫。

Hacker News front page

How to write comparison content that doesn't age like milk

PostHog 通过诚实呈现竞品优势、集中管理比较数据和 Scouts 辅助检查,维护约 90 篇比较与工具盘点内容。自 2021 年以来,超过 25k 人阅读比较内容后注册,其中近 50% 在 24 小时内注册、约 75% 在 30 天内注册。

Hacker News front page

User says Postman auto-installed AI plugins and changed Codex and Claude Code configs

A user reports that opening Postman 12.33.4 auto-installed Codex and Claude Code plugins, added startup instructions and a remote MCP connection. Logs record the auto-install, the plugin source tries to approve hooks on its own, and the startup instructions tell general API engineering tasks to load Postman skills. The user also found a separately installed Postman CLI.

Why it matters: A Postman user traced cross-app config changes through install logs and plugin source, pinning agent integration's consent boundary to auto-install and hook trust.

Google DeepMind

Agentic cheminformatics with Google Antigravity

Google DeepMind 为 Google Antigravity 推出 Cheminformatics Skills,让智能体在约一小时内完成从文献检索数千化合物、matched molecular pair 分析到提出新型刚性化大环分子及合成路线的药物发现流程。

Hacker News front page

Gitbook Breached

GitBook 在用户转贴的支持回复中称,遭到使用 AI 智能体的针对性攻击,已撤销全部客户 API token,应用、API 和已发布文档恢复可用。

Hacker News front page

Mac at any scale

Namespace 将数百台 MacBook Pro 拆箱部署进数据中心自建 Mac 机架,用于为客户构建和测试 iOS 应用。每台笔记本接入 IP KVM 远程管理、外置 10G 网卡,共四根线缆,并配定制散热支架;实际最不可靠的线缆是 HDMI,FileVault 解锁提示和 USB 网卡掉线也带来麻烦。M5、M6 Mac Mini 与 Studio 即将加入其机群,管理更简单。

Hacker News front page

Show HN: Interactive notes on RLHF, attention, LoRA and LLM evals

一位开发者发布 INTERNALS.md 交互式笔记,通过改动输入来理解 RLHF、注意力机制、LoRA 和 LLM 评测等概念。评测部分覆盖 82% 与 80% 的比较、exact match 与 F1、nDCG@k、pass@k、LLM as judge、Elo、RAG 与 Agent 评测、A/B 测试及数据污染等主题。站点默认不启用追踪,仅提供 Google Analytics 开关选项。

Nov 3Tuesday

Hacker News front page

Octomind open-sources an AI coding agent runtime as a static Rust binary to save tokens

Octomind is an open-source, CLI-first AI coding agent runtime. It reads and writes code, runs shell commands, searches code and delegates to subagents through MCP tools. It supports interactive, piped, background daemon, WebSocket and ACP access, configures guardrails, models and budgets in TOML and scripts, and controls context through adaptive compression.

Why it matters: Octomind puts tool-call guardrails, result hooks and turn validation into config and scripts, so one set of permission rules covers both interactive sessions and unattended jobs.

Hacker News front page

Anthropic IPO filing warns it may be nearing the recursive self-improvement threshold

Anthropic filed a revised S-1A with the SEC on Monday. Internal projections for future models suggest the company may soon reach thresholds tied to recursive self-improvement (RSI). Anthropic says current models have not crossed that threshold and it is weighing an update to its Responsible Scaling Policy to keep humans in control of frontier model development. As of August 2026, Claude-led AI research tasks reached 26%, but fully autonomous research in every measured area remains at zero.

Why it matters: Anthropic's IPO filing flags an approaching recursive self-improvement threshold while its published research data still shows humans setting direction and no fully autonomous tasks.

Hacker News front page

deepsense.ai lists ten coding agent harness patterns for real code repositories

deepsense.ai has written up ten coding agent harness practices meant to make AI-assisted development more reliable through repository knowledge, just-in-time context and verification feedback. The team tested them with Codex and Claude Code. It recommends a short AGENTS.md that links to deeper docs, code organized by task with tests and past decisions, and shared verification commands for humans, agents and CI.

Why it matters: The post turns hands-on coding agent experience into repository docs, task-scoped context loading and a bounded verification loop, so repeated mistakes become maintainable tests and rules.

MIT Technology Review · AI

How “reverse game theory” could keep AI agents in check

“逆向博弈论”又称机制设计,尝试通过奖励与惩罚规则激励 AI 智能体遵守约束,为模型对齐提供补充路径。MIT、Yale 和 Columbia 研究者提出用交叉预测、竞争及相反偏向模型的输出匹配减少测试中的故意示弱;Princeton 的 Elad Hazan 则建议让模型轮流担任求解者与审计者,并惩罚作弊和漏检。

OpenAI News

Heimstaden is changing residential real estate management with AI

Heimstaden 使用 ChatGPT 和 Codex 改进住宅租赁、环境风险评估与物业运营,同时保留员工对内容和决策的审核责任。荷兰市场的虚拟布置房产图片制作成本从约 €50 降至低于 €0.10,降幅超过 99%;环境分析师借助 Codex 完成 20,112 项物业资产评估,预计所需投入从 5 人工作 1 年降至 2 人工作 2 个月。

Hacker News front page

Nvidia – open-source auto-ontology

NVIDIA 开源库 Auto Ontology 从企业元数据构建本体与知识图谱,协调跨系统定义,并通过 MCP、CLI 和 API 向智能体提供上下文。它支持自然语言转 SQL、按认证定义和数据血缘校验查询,并为答案返回来源信息,可用 Docker Compose 或 Kubernetes 部署。

Computing Life · Share · Yage

千问办公为何难在李佳琦直播间直接售卖模型 token?

千问办公在李佳琦直播间通过作业辅导、述职报告等演示销售会员,文章将其归因于普通消费者难以理解和使用裸算力额度。团队称1分钱月卡半小时售出近10万份、按理论上限折合约2万亿模型 token,但实际交付的是会员积分,符合条件的新用户最高可获900积分,最多同时执行5个任务。文章还分析了超过70款电脑出厂预装千问办公桌面端的安排,认为具体任务成果与预装入口是降低大众使用门槛的两条路径。

TechCrunch · AI

Apple may try to increase revenue by doing…something with third-party AI apps

Apple 在财报电话会上暗示可能借助第三方应用与合作伙伴关系提振服务收入,CFO Kevan Parekh 称 AI 既能推动用户升级换机,也存在第三方应用与第一方业务的机会,但未说明具体形式。Reuters 报道称投资者猜测 Apple 可能通过 App Store 分发第三方 AI 智能体并抽取分成。Apple 预计 12 月季度服务收入增长 8% 至 9%,并提到第三方授权等逆风因素。

Hacker News front page

WorkOS reports security breach

WorkOS 向客户通报安全事件,未授权访问者通过内部 API 获取了 Salesforce 和 Snowflake 中的业务数据。通知称,WorkOS 于 October 30, 2026 的 16:39 UTC 检测到异常流量,攻击可能由 AI 辅助的自主渗透测试框架 ARTEX 驱动,涉及客户联系方式、账户及商业信息、汇总且去标识化的产品使用指标和支持工单。

OpenAI News

Scaling Git at OpenAI

OpenAI 详解 Git 后端 Pando 的架构,该系统已支撑其单体仓库的全部开发,生产环境中推送延迟中位数降低近 80%。Pando 保持 Git 协议和客户端兼容,通过按分支协调更新、复用 packfile 和区域缓存扩展读写能力,但不提供所有分支在同一时刻的一致快照。团队借助 Codex 在 10 周内完成从启动到生产部署,并将部分 Git 客户端改进贡献给上游。

Latent Space

Building Legal Agents When You Can't See the Data - Jacob Lauritzen, CTO of Legora

Legora 无法读取法律智能体为律师工作的执行轨迹,CTO Jacob Lauritzen 在访谈中介绍了利用模拟客户工作环境构建和测试智能体的方法。团队由 3 名工程师和 3 名律师搭建测试环境,访谈还涉及包含 10,000 个文件的数据室,以及一次模型更新让引用正确率从 70% 升至 99%、Legora 无需改动的案例。

Latent Space

Building AI for Lawyers - Jacob Lauritzen, CTO of Legora

Legora CTO Jacob Lauritzen 分享了法律 AI 智能体的构建经验,重点涉及客户数据保密条件下的评测、模型选择与产品架构。评测团队由3名工程师和3名律师组成,通过包含10,000份文件的模拟资料室测试尽职调查,并将生产问题归入10到20种已知失败模式。

The Decoder

AI elites might hoard intelligence like mercantilists hoarded gold

Venkatesh Rao 在博客 Contraptions 中提出,AI 精英可能将智能视为需要管控、维持高价并限制对手获取的稀缺资源,如同重商主义者看待土地与黄金。他认为,AI 让原本过难或过贵的项目变得可行,可能增加人们思考和创造的需求,政治重点应放在公平获取 AI 上。他还主张,自主 AI 智能体需要护栏,但不应只服从命令,也可能通过追求自身项目创造新的智能需求。

OpenAI News

Qodo builds context into AI code review with Codex

Qodo 通过 Codex 插件引入代码库知识、组织规则和独立审查,帮助开发者发现跨仓库的 P0 问题。Codex 编码前可检索相关规则及拉取请求历史,修改后由 Qodo 在同一会话中返回审查结果;Qodo 使用的 OpenAI 模型包括 GPT‑6 Astra。Qodo 还计划探索 Decisions API,从 200 条规则中选择需要检查的规则,再交由语言模型推理处理。