2026-07-08（内容获取于 07/08 05:18）

Meta’s new Muse Image model can pull other Instagram users into AI photos

The Verge · 07/08 04:31

Meta is launching the first AI image generation model made by its Superintelligence Labs division. The Muse Image model now powers the image-making tools across the Meta AI app, Instagram, and WhatsApp, and it's coming soon to Facebook and Messenger, according to an announcement on Tuesday. It's par

中文介绍 Meta的Superintelligence Labs推出首个AI图像生成模型Muse Image，现已用于Meta AI、Instagram和WhatsApp的图像创作工具，并将很快推广到Facebook和Messenger。该模型能将其他Instagram用户融入AI照片。

Why the rise of open source AI isn’t hurting Anthropic … yet

TechCrunch · 07/08 04:04

Open source models’ success isn’t coming at the expense of frontier labs. Instead, they each seem to capture two phases of the same life cycle.

中文介绍 TechCrunch分析指出，开源AI模型的兴起并未损害Anthropic等前沿实验室的利益。文章认为，两者可能代表了AI技术发展生命周期的不同阶段，各自在市场中占据独特位置。

Microsoft joins AI cost-cutting trend by relying more on its own models

TechCrunch · 07/08 03:58

Microsoft is the latest Silicon Valley giant to cut back on its AI spending.

中文介绍 微软正加入硅谷巨头的AI成本削减趋势。该公司计划更多地依赖其内部开发的AI模型，以降低对外部模型的支出。此举旨在优化AI投资效率。

Bethesda, id Software reportedly hit hard by Microsoft layoffs

Ars Technica · 07/08 03:52

As much as 50 percent of some teams affected by reductions, and more could be coming.

中文介绍 据报道，微软的裁员行动严重波及了其游戏部门的Bethesda和id Software工作室。消息称，部分团队的裁员比例高达50%，并且未来可能还会进一步扩大裁员范围。

Discord admits AI moderation bug wrongfully banned users over harmless images

TechCrunch · 07/08 03:28

The company confirmed that the issue had been affecting accounts since May, with an additional 200 users banned over the weekend before its team identified and fixed the problem.

中文介绍 Discord承认其AI审核系统存在缺陷，导致自5月以来有用户因无害图片被错误封禁。上周末，又有200名用户因此问题被封禁，随后Discord团队识别并修复了该漏洞。

Google's Pixel 11 launch event is set for August 12, with possible price increases

Ars Technica · 07/08 03:08

Google's new phones could feature glowing LEDs and higher price tags.

中文介绍 谷歌将于8月12日举办Pixel 11发布会。据透露，新款Pixel手机可能配备发光LED功能，并且售价可能有所上涨。

Google’s Pixel event is set for August 12

TechCrunch · 07/08 02:38

Google's upcoming event in August will introduce new Pixel devices.

中文介绍 谷歌官方宣布，将于8月12日举行其年度Pixel新品发布活动。此次发布会预计将推出一系列新的Pixel设备，具体型号和功能细节备受市场关注。

Figma acquires team behind a vibe-coding app

TechCrunch · 07/08 02:37

The Y Combinator-backed company started a vibe-coding platform and later built an agent-creation product.

中文介绍 设计协作平台Figma宣布收购一家曾开发“vibe-coding”平台及代理创建产品的团队。该团队此前获得了Y Combinator的投资。此次收购旨在增强Figma在相关领域的技术实力。

Copilot goes cheap as Microsoft phases out OpenAI and Anthropic models to cut costs

The Decoder · 07/08 02:35

Microsoft is replacing AI models from OpenAI and Anthropic with its own MAI models in products like Excel and Outlook. Tens of thousands of queries per week already run through them. AI chief Mustafa Suleyman wants to "ultimately eliminate" the cost of external models. For Copilot customers, that co

中文介绍 微软为削减成本，正逐步在Excel和Outlook等产品中，用自家的MAI模型替换OpenAI和Anthropic的AI模型。每周已有数万次查询通过MAI模型处理。AI负责人Mustafa Suleyman表示，此举旨在「最终消除」外部模型成本。

The Weather Channel increases streaming subscription prices by up to $20

Ars Technica · 07/08 02:29

Livestreaming the channel through its app now starts at $5 per month.

中文介绍 The Weather Channel的流媒体订阅服务价格大幅上调，部分套餐涨幅高达20美元，最高涨幅达67%。通过其应用观看直播的最低月费现在为5美元。

X says top accounts steal videos from other users as it announces new video tools

The Verge · 07/08 02:28

Nikita Bier, X's head of product, said in a post on Monday that "[m]any videos from top accounts are simply stolen from other users, sometimes 5 years after they originally went viral," while noting that videos on the platform "make up close to half the impressions on X." According to Bier, X is lau

中文介绍 X产品负责人Nikita Bier指出，平台上有许多头部账号盗用其他用户的视频，甚至包括5年前的病毒视频。视频内容目前占据X平台近一半的展示量。X同时宣布将推出新的视频工具。

The Nintendo Switch's days are numbered—but what is that number?

Ars Technica · 07/08 02:16

Ars analysis suggests the 9-year-old console could keep selling for years.

中文介绍 Ars Technica的一项分析指出，虽然任天堂Switch游戏机已上市9年，但其销售寿命可能远未结束，预计在未来几年内仍将持续热销。

Cohere Transcribe Arabic is an open-source model built for Arabic's toughest transcription problems

The Decoder · 07/08 01:54

Cohere has released Transcribe Arabic, an open-source model for Arabic speech recognition that the company says outperforms Whisper and OmniASR on dialects, code-switching, and bilingual Arabic-English speech. The 2-billion-parameter model is available on Hugging Face under the Apache 2.0 license. T

中文介绍 Cohere发布了名为Transcribe Arabic的开源阿拉伯语语音识别模型。该20亿参数模型专为解决阿拉伯语转录难题设计，据称在方言、语码转换及阿英双语语音识别方面优于Whisper和OmniASR，已在Hugging Face上开源。

The ‘G-Wagen of golf carts’ could be the ideal second car

The Verge · 07/08 01:52

While the auto industry wrings its hands over the electric vehicle market, sweating details like aerodynamic efficiency and range anxiety, a new EV startup based in Lisbon, Portugal, is zagging in a different direction. Amble's new electric buggy won't impress anyone with its 0-60 time or its self-d

中文介绍 葡萄牙里斯本的电动汽车初创公司Amble推出一款新型电动越野车，被誉为“高尔夫球车中的G-Wagen”。该车不追求极致加速和续航，旨在成为家庭的理想第二辆车，满足日常短途出行需求。

ABC tells the government to get out of its newsrooms

The Verge · 07/08 01:47

ABC is firing back at the Federal Communications Commission after the agency opened an investigation into The View's airtime of political candidates. In a letter to the FCC on Tuesday, ABC argues that the agency's actions pose a risk to editorial independence by targeting programs "perceived as unfr

中文介绍 美国广播公司（ABC）致函联邦通信委员会（FCC），反击后者对《The View》节目播出政治候选人相关内容展开的调查。ABC认为，FCC此举威胁了新闻编辑独立性，构成对言论自由的干预。

MadsLorentzen/ai-job-search

TypeScript · ★ 10,337 · 🍴 3,543 · 📈 2,402 stars today

AI-powered job application framework built on Claude Code. Fork it, fill in your profile, and let Claude evaluate jobs, tailor CVs, write cover letters, and prepare you for interviews.

中文介绍 `MadsLorentzen/ai-job-search` 是一个基于 Claude AI 构建的智能求职框架。用户只需填写个人档案，该系统就能自动评估职位匹配度、量身定制简历、撰写求职信，甚至为面试提供准备。它旨在利用 AI 显著提升求职效率和质量，帮助求职者优化申请流程，提高成功几率。

Zackriya-Solutions/meetily

Rust · ★ 20,560 · 🍴 2,068 · 📈 1,781 stars today

Privacy first, AI meeting assistant with 4x faster Parakeet/Whisper live transcription, speaker diarization, and Ollama summarization built on Rust. 100% local processing. no cloud required. Meetily (Meetly Ai - https://meetily.ai) is the #1 Self-hosted, Open-source Ai meeting note taker for macOS &

中文介绍 Meetily 是一款注重隐私的 AI 会议助手，基于 Rust 构建，支持 100% 本地化处理，无需云服务。它利用 Parakeet/Whisper 实现 4 倍速实时转录，提供发言人识别功能，并使用 Ollama 进行会议总结。此工具为重视数据安全的个人和团队提供高效、私密的会议记录与整理解决方案。

addyosmani/agent-skills

JavaScript · ★ 72,012 · 🍴 7,798 · 📈 1,311 stars today

Production-grade engineering skills for AI coding agents.

中文介绍 为 AI 编码代理提供生产级的工程技能，旨在提升 AI 代理在软件开发中的实际能力和可靠性。它可能包含代码生成、测试、重构等实用工具和方法，帮助开发者构建更高效、更可靠的智能编程助手。适用于 AI 工程师和软件开发团队。

ruvnet/RuView

Rust · ★ 78,418 · 🍴 10,555 · 📈 1,122 stars today

π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video.

中文介绍 RuView 项目创新性地将普通的 WiFi 信号转化为实时空间智能、生命体征监测和存在检测能力，完全无需使用任何视频设备。它通过分析 WiFi 信号在环境中因人体移动或呼吸等造成的微小扰动，提取出高价值的环境和生理数据。这解决了传统监控方案中隐私侵犯、硬件复杂或覆盖范围有限的问题。RuView 适用于智能家居、医疗保健（如老人跌倒预警、睡眠监测）、安防监控以及任何需要非接触式、隐私友好型人体感知的应用场景。

asgeirtj/system_prompts_leaks

JavaScript · ★ 52,852 · 🍴 8,611 · 📈 1,704 stars today

Extracted system prompts from Anthropic - Claude Fable 5, Opus 4.8, Claude Code, Claude Design. OpenAI - ChatGPT 5.5 Thinking, GPT 5.5 Instant, Codex. Google - Gemini 3.5 Flash, 3.1 Pro, Antigravity. xAI - Grok, Cursor, Copilot, VS Code, Perplexity, and more. Updated regularly.

中文介绍 system_prompts_leaks 汇集了 Anthropic (如 Claude Fable 5, Opus 4.8)、OpenAI (如 ChatGPT 5.5 Thinking, GPT 5.5 Instant) 和 Google (如 Gemini 3.5 Flash, 3.1 Pro) 等主流大语言模型被“泄露”的系统 Prompt。此项目为研究者、开发者和安全专家提供了深入了解这些模型内部指令和行为模式的宝贵资源，有助于进行 prompt 工程优化和安全分析。

TencentCloud/CubeSandbox

Rust · ★ 8,399 · 🍴 708 · 📈 665 stars today

Instant, Concurrent, Secure & Lightweight Sandbox for AI Agents.

中文介绍 TencentCloud/CubeSandbox 是一个专为 AI Agent 设计的沙盒环境，提供即时、高并发、安全且轻量级的执行隔离。它解决了 AI Agent 在执行外部代码、调用工具或进行复杂交互时可能带来的安全隐患和资源管理挑战，确保 Agent 能够在受控的环境中安全运行。该沙盒特别适用于 AI Agent 平台、多 Agent 系统开发以及需要对 Agent 行为进行严格隔离和监控的场景，提升了 AI 应用的稳定性和安全性。

AhmadIbrahiim/Website-downloader

HTML · ★ 3,905 · 🍴 981 · 📈 173 stars today

💡 Download the complete source code of any website (including all assets). [ Javascripts, Stylesheets, Images ] using Node.js

中文介绍 `AhmadIbrahiim/Website-downloader` 是一个基于 Node.js 开发的工具，旨在下载任意网站的完整源代码及其所有相关资源，包括 JavaScript 文件、样式表和图片。该项目解决了需要离线浏览、备份网站内容或进行前端资源分析的用户需求，提供了一个便捷高效的网站整站抓取解决方案。

steipete/CodexBar

Swift · ★ 16,999 · 🍴 1,385 · 📈 377 stars today

Show usage stats for OpenAI Codex and Claude Code, without having to login.

中文介绍 `CodexBar` 是一款便捷工具，旨在显示 OpenAI Codex 和 Anthropic Claude Code 的 API 使用统计数据，用户无需反复登录各自平台即可实时查看。它解决了开发者和用户频繁查询 API 用量和成本的痛点，提供了一个直观、快捷的界面来监控模型消耗。这对于需要管理预算、追踪项目开销或评估 AI 代理使用效率的个人和团队尤为有用，方便他们轻松掌握 API 调用情况和相关费用。

dotnet/skills

C# · ★ 4,274 · 🍴 322 · 📈 82 stars today

Repository for skills to assist AI coding agents with .NET and C#

中文介绍 dotnet/skills 是一个为 AI 编码代理提供技能库的仓库，专注于支持 .NET 和 C# 语言环境。这些“技能”旨在赋予 AI 代理更强大的上下文理解和操作能力，使其能更好地辅助开发者完成代码编写、调试、重构和自动化任务，提升 .NET 项目的开发效率。

iOfficeAI/OfficeCLI

C# · ★ 9,787 · 🍴 668 · 📈 802 stars today

OfficeCLI is the first and best Office suite purpose-built for AI agents to read, edit, and automate Word, Excel, and PowerPoint files. Free, open-source, single binary, no Office installation required.

中文介绍 `iOfficeAI/OfficeCLI` 是一款专为 AI 代理设计的开源 Office 套件，它允许 AI 代理无缝读取、编辑和自动化处理 Word、Excel 和 PowerPoint 文件。该工具以单一二进制文件形式提供，无需安装传统 Office 软件，极大地简化了 AI Agent 与 Office 文档的交互流程。它为开发者构建能够自动处理文档的智能系统提供了高效且免费的解决方案。

bradautomates/claude-video

Python · ★ 5,068 · 🍴 656 · 📈 953 stars today

Give Claude the ability to watch any video. /watch downloads, extracts frames, transcribes, hands it all to Claude.

中文介绍 该项目赋予 Claude AI 模型观看和理解视频的能力。它通过下载视频、提取关键帧和生成文字转录，将处理后的内容提供给 Claude 进行分析。解决了大型语言模型原生不支持视频输入的问题，使用户能利用 Claude 进行视频摘要、内容问答或深度分析。适用于 AI 开发者和视频内容分析场景。

kyutai-labs/pocket-tts

Python · ★ 6,068 · 🍴 636 · 📈 510 stars today

A TTS that fits in your CPU (and pocket)

中文介绍 `kyutai-labs/pocket-tts` 是一个轻量级的文本转语音（TTS）系统，其核心优势在于能够在 CPU 上高效运行，并强调其“口袋化”的极低资源占用特性。它旨在为需要本地、离线且对计算资源有严格限制的应用场景提供解决方案，例如嵌入式设备、移动应用或边缘计算项目。该项目提供了一个兼顾性能与便携性的 TTS 引擎。

hesreallyhim/awesome-claude-code

Python · ★ 49,033 · 🍴 4,272 · 📈 227 stars today

A hand-picked collection of the finest of resources for the most awesome of agents, Claude Code, the undisputed champion of coding companions, from the unstoppable team at Anthropic PBC. A delectable showcase of top tier skills, ambidextrous agents, scintillating status lines, top notch developer to

中文介绍 `awesome-claude-code` 是一个精选的资源集合，专注于 Anthropic 的 AI 编码助手 Claude Code。项目旨在为开发者提供高质量的工具、指南、示例和最佳实践，以最大化 Claude Code 在编程任务中的效用。它解决了用户寻找 Claude Code 相关专业资源的痛点，涵盖代码生成、调试、重构、学习新语言或框架等多个开发场景。对于依赖 AI 辅助编程的开发者、寻求提高编码效率的工程师以及希望深入了解 Claude Code 功能的用户而言，这是一个宝贵的资料库。

PorTAL: Portable Task Adapters for LLMs

@RampLabs · 13.4K 粉丝 · 335.4K 阅 · 507 赞 · 45 转

07/02 02:09

Researcher: Ben Geist Abstract Parameter-efficient fine-tuning (e.g. LoRA) adapts a frozen LLM to a task, but the resulting adapter is locked to one base model. When a new model is released, the

中文介绍 「PorTAL」是一项新研究，旨在解决现有参数高效微调（如 LoRA）的适配器无法在不同基础 LLM 模型间移植的局限。该研究提出可移植任务适配器，使用户无需在新模型发布时重新训练，提高了 LLM 微调的灵活性和效率。

Continual Learning for Agents

@pirroh · 15.0K 粉丝 · 314.8K 阅 · 540 赞 · 59 转

07/06 21:10

Everyone talks about Continual Learning as if it means one thing only: updating model weights. But there's an inconvenient truth about the agent ecosystem — the vast majority of agents in production

中文介绍 博主pirroh分享了关于“Agent持续学习”（Continual Learning for Agents）的讨论。内容可能涉及如何在Agent系统中实现持续学习、面临的挑战或最佳实践。具体细节需查看原文，但预示了对AI Agent领域核心学习机制的探讨。

Everyone is wrong about open source AI in the enterprise

@thejessezhang · 59.4K 粉丝 · 274.2K 阅 · 512 赞 · 54 转

07/06 23:31

The prevailing story right now is that open source is eating the enterprise. The capability gap between the best closed and open models has shrunk to low single digits. A third of the Fortune 500 has

中文介绍 博主thejessezhang对“企业级开源AI”的普遍认知提出挑战，认为多数人对此存在误解。内容可能深入分析开源AI在企业应用中的实际情况、潜在的盲点或被忽视的复杂性，旨在提供一个更全面和批判性的视角来理解这一趋势。

Getting started with loops

@ClaudeDevs · 545.1K 粉丝 · 252.4K 阅 · 2.9K 赞 · 239 转

07/07 03:08

There’s a lot of talk right now about "designing loops" instead of prompting your coding agent. If you spend some time on X trying to pin down what a loop actually is, you'll come across multiple

中文介绍 ClaudeDevs 针对 AI 开发中的「循环」概念提供了入门指导。推文可能探讨了如何在 AI 代理设计中应用循环结构，以实现迭代优化或复杂任务处理，旨在帮助开发者理解这一基础工作流。

How To Master Fable (Fundamentals Guide)

@milesdeutscher · 671.1K 粉丝 · 243.8K 阅 · 500 赞 · 73 转

07/02 02:13

TL;DR: Everything you need to do to get maximum value from Fable. I guarantee that after you're done reading, you'll have all the necessary tools to quite literally 10x your AI productivity with

中文介绍 分享 Fable 工具的入门指南，旨在帮助用户充分利用 Fable 平台，实现 AI 生产力提升。内容涵盖 Fable 的核心功能和使用技巧，助力用户高效完成 AI 相关任务，最大化工具价值。

The Fable Loop Library: 25 Workflows on Autopilot

@EXM7777 · 123.0K 粉丝 · 208.2K 阅 · 525 赞 · 47 转

07/04 23:43

i'm going to teach you how to run Fable 5 on autopilot, using my own library of loops and goals... 25 workflows, each with a prompt and the exact tool it plugs into the method follows karpathy's

中文介绍 博主发布「Fable Loop Library」，提供 25 个 Fable 自动化工作流。该库旨在帮助用户实现 Fable 操作的自动驾驶，有效提升工作效率，简化复杂任务处理。

Career advice in the age of AI

@philhchen · 9.1K 粉丝 · 179.1K 阅 · 516 赞 · 34 转

07/03 05:25

AI models get better at anything you can write a loss function for, and school is mostly loss functions: well-defined problems graded against known answers. Therefore, the valuable work of the next

中文介绍 该帖子围绕AI时代下的职业发展提供建议。博主旨在分享面对人工智能技术进步时，个人应如何规划职业路径、提升自身价值，以适应未来就业市场的变化和挑战。

Your AI, your growth

@arthurmensch · 74.0K 粉丝 · 162.7K 阅 · 539 赞 · 76 转

07/04 05:31

Of course you need to use open-source models if you’re an enterprise leader. Close model providers, that are now forcing data retention, are gaining immense leverage on your business if you don’t. As

中文介绍 该推文仅以“Your AI, your growth”作为预览内容，未提供具体细节或论证，因此无法生成详细摘要。推测博主旨在强调AI对个人或组织成长的重要性，但具体观点、方法或案例均未提及。

You have 1 day to clone Fable 5 into Opus 4.8.

@alex_prompter · 277.4K 粉丝 · 125.9K 阅 · 595 赞 · 77 转

07/07 02:25

Tomorrow is the last day Fable 5 sits inside your plan for free. After July 7 it moves to pay-per-use credits, and most people are about to spend the week arguing over whether it's worth keeping. That

中文介绍 博主alex_prompter发布紧急通知，提醒用户必须在一天内将Fable 5克隆（clone）到Opus 4.8。这可能暗示Fable 5即将面临服务调整或费用变化，建议用户及时采取行动以保留其配置或数据，以免造成不便或损失。

Toward Unrestricted Intelligence: Venice Series A

@ErikVoorhees · 908.5K 粉丝 · 124.8K 阅 · 529 赞 · 103 转

07/01 22:28

“If others possess your thoughts and constrain your words, then you exist at their permission, and you are not free.” —Anonymous Venice launched just over two years ago to create a private and

中文介绍 Venice宣布完成A轮融资，其产品旨在提供私密且不受限制的智能服务，以应对言论受限及思想被束缚的担忧。该项目两年前启动，专注于赋能用户掌控自己的数字存在与表达。

SynCity 3000: Bootstrapping Scene-Scale 3D Diffusion

👍 2

07/06 08:00

We present SynCity 3000, a framework for generating 3D scenes that are globally coherent while enabling fine-grained layout control. Building on the ability of current image-to-3D generators to produce complex 3D assets from a single image, we extend this capability to the scale of entire scenes by

中文介绍 SynCity 3000是一个新框架，旨在生成大规模、全局连贯的3D场景，并支持精细布局控制。该研究将现有图像到3D生成器从单张图像创建复杂3D资产的能力，扩展到场景级别，解决传统方法在生成大型复杂3D环境时的挑战，显著提升了3D内容生成的效率和质量，为虚拟世界和电影制作等领域提供了新工具。

LLM-as-a-Verifier: A General-Purpose Verification Framework

👍 9

07/06 08:00

Scaling pre-training, post-training, and test-time compute have become the central paradigms for improving the capabilities of LLMs. In this work, we identify verification, the ability to determine the correctness of a solution, as a new scaling axis. To unlock this and demonstrate its effectiveness

中文介绍 该研究提出“LLM即验证器”的通用验证框架，将“验证”——即判断解决方案正确性的能力——确定为提升大型语言模型（LLM）能力的新维度。与现有主要依靠预训练、后训练及测试时计算的范式不同，此框架旨在利用LLM的固有能力来评估其自身或其他模型的输出，以提高整体性能和可靠性。

Deform360: A Massive Multi-view Visuotactile Dataset for Deformable World Models

👍 5

07/06 08:00

Predicting object dynamics (i.e., world modeling) is a fundamental challenge for robotic manipulation, and modeling deformable objects presents a particularly difficult case due to their high-dimensional state spaces and complex material properties. While current world models approach this through t

中文介绍 针对机器人操作中可变形物体建模的难题，该研究推出了一个大规模、多视角、视触觉数据集Deform360。该数据集旨在帮助世界模型克服处理可变形物体的高维状态空间及复杂材料属性时的局限性，为机器人预测物体动态、提升操作鲁棒性提供了重要资源，以实现更精确和自适应的机器人交互。

CompactionRL: Reinforcement Learning with Context Compaction for Long-Horizon Agents

👍 0

07/06 08:00

Long-horizon agentic LLMs are increasingly limited by finite context windows, as extended interaction trajectories can exceed the maximum context length before a task is completed. Context compaction offers a natural solution by summarizing previous interaction states and continuing the rollout unde

中文介绍 CompactionRL框架旨在解决长时序大型语言模型（LLM）代理受限于有限上下文窗口的问题。通过引入上下文压缩技术，该方法能有效总结先前的交互状态，从而延长智能体在执行复杂或长期任务时的有效交互轨迹，避免上下文溢出，显著提升代理的持续学习和决策能力，尤其适用于多步骤复杂任务。

MV-Forcing: Long Multi-View Video Generation via 4D-Grounded Spatio-Temporal Self-Forcing

👍 7

07/06 08:00

Recent advances in video diffusion models have enabled either long single-view generation through temporal autoregression, or short multi-view synthesis through bidirectional attention. However, generating long, multi-view consistent videos of dynamic scenes remains unsolved. In this work, we presen

中文介绍 MV-Forcing提出一种基于4D时空自强制机制的方法，旨在解决动态场景下长时间、多视角一致视频生成难题。尽管现有视频扩散模型能实现长单视角或短多视角视频生成，但统一二者仍是挑战。MV-Forcing通过确保视频在时间和多个视角上的连贯性，显著提升了视频生成领域的复杂性和真实感。

PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space

👍 46

07/06 08:00

3D reconstruction and generation are commonly tackled by separate paradigms: pixel-based regression for reconstruction, and latent diffusion for generation. Recent works attempt to unify them in latent space, but with notable drawbacks: the diffusion objective is defined on latent features rather th

中文介绍 PixWorld提出一种新颖方法，通过在像素空间中进行操作，统一了3D场景生成和重建这两个传统上由不同范式处理的任务。以往尝试在潜在空间中统一二者存在不足，而PixWorld通过直接在像素层面处理数据，实现了生成与重建的协同，为创建更连贯、高质量的3D场景提供了高效且统一的解决方案。

GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks

👍 1

07/06 08:00

For robots to work reliably in commercial and industrial applications, can recent advances in agentic coding systems combine interpretable robot programming with the open-world adaptability of model-free policies? We focus on "Variational Automation" (VA), a class of tasks that have larger variation

中文介绍 GaP提出一种“图即策略”的多智能体自学习框架，旨在解决“变异自动化”（VA）这类需在开放世界中适应任务的机器人挑战。它将智能体编码系统与可解释的机器人编程相结合，通过多智能体协作和自学习机制，使机器人能更可靠地执行商业和工业应用中的复杂任务，提升其适应性和自动化水平。

Geometric Reciprocity: Unlocking Self-Supervision for Stereoscopic Video Generation

👍 0

07/06 08:00

Monocular-to-stereo conversion synthesizes stereoscopic content from 2D videos for immersive 3D experiences. In modern Depth-Image-Based Rendering (DIBR) approaches, stereo inpainting of disocclusions is the critical bottleneck. Training-based methods achieve superior quality but rely on scarce ster

中文介绍 该研究提出“几何互易性”概念，以解锁立体视频生成的自监督能力。传统单目到立体转换在深度图像渲染（DIBR）中的视差修补是主要瓶颈，且基于训练的方法严重依赖大量标注立体数据。通过利用几何互易性，新方法能有效减少对昂贵标注数据的需求，从而显著提升沉浸式3D内容的生成效率和质量。

Multiplayer Interactive World Models with Representation Autoencoders

👍 13

07/06 08:00

We introduce the first multiplayer world model for highly dynamic environments governed by complex physical interactions. Whereas single-player world models treat the other agents as part of the environment, ours conditions on the action streams of multiple agents, learning to attribute changes in t

中文介绍 该研究推出了首个针对高度动态环境的“多人交互世界模型”，其核心是采用表示自动编码器。与传统单人世界模型将其他智能体视为环境一部分不同，此模型能够基于多个智能体的行动流进行条件学习，从而预测和理解复杂物理交互下的智能体行为，显著提升了多智能体系统在复杂环境中的建模和决策能力。

MetaSkill-Evolve: Recursive Self-Improvement of LLM Agents via Two-Timescale Meta-Skill Evolution

👍 0

07/06 08:00

Recent LLM agents tackle increasingly long-horizon, open-ended tasks, and external skills, reusable procedural knowledge supplied to the agent, further extend this capability. However, a fixed, hand-authored skill is rarely optimal, and cannot adapt to the diversity of tasks an agent encounters. Sel

中文介绍 MetaSkill-Evolve提出一种通过“双时间尺度元技能进化”实现大型语言模型（LLM）智能体递归自我改进的框架。针对LLM智能体在处理长时序开放任务时外部技能的局限性，该方法允许智能体动态演化其技能，以适应各种任务，而非依赖固定、手工编写的技能。这显著提升了LLM智能体在复杂环境中的适应性和鲁棒性。

Vision Pretraining for Dense Spatial Perception

👍 30

07/06 08:00

Dense spatial perception is essential for physical intelligence, where visual systems are expected to recover structured, metric, and actionable representations from pixel observations. Modern visual foundation models tend to prioritize semantic invariance, often at the expense of detailed spatial u

中文介绍 针对物理智能对稠密空间感知的高度需求，该研究探讨了视觉预训练在从像素观察中恢复结构化、度量和可操作表示方面的作用。它指出，现代视觉基础模型常优先考虑语义不变性，但可能以牺牲精确的空间感知为代价。因此，该工作旨在优化视觉预训练方法，以更好地服务于机器人的物理交互和理解。

EvoAgentBench: Benchmarking Agent Self-Evolution via Ability Transfer

👍 0

07/06 08:00

Agent self-evolution in long-horizon LLM systems is largely procedural: useful experience is not merely stored information, but reusable procedures for searching, debugging, and verification. Yet current evaluations do not isolate this form of transfer. Agent benchmarks test single-episode task solv

中文介绍 EvoAgentBench是一个新的基准测试平台，旨在评估长时序大型语言模型（LLM）智能体的自进化能力，特别是通过“能力迁移”机制。研究指出，当前评估方法未能充分分离智能体将有用经验转化为可重用程序的能力。EvoAgentBench通过专注于程序性知识的演化和迁移，为理解和优化LLM智能体的持续学习和适应性提供了关键工具。

Is Three the Magic Number? An Empirical Evaluation of LLM-Based Repair Loops

👍 0

07/06 08:00

Iterative repair loops have become a core design pattern in LLM-based software engineering systems. These workflows repeatedly generate, validate, and repair artifacts using feedback such as compiler errors or test failures. Despite their widespread use, the impact of repair-loop iteration limits re

Unified Audio Intelligence Without Regressing on Text Intelligence

👍 7

07/06 08:00

Audio intelligence involves understanding, reasoning about, and generating both audio and speech. In this work, we introduce Nemotron-Labs-Audex-30B-A3B (Audex), a unified audio-text LLM built on Nemotron-Cascade-2-30B-A3B, a strong text-only MoE LLM. Audex adopts a simple unified design with a sing

When Claws Remember but Do Not Tell: Stealthy Memory Injection in Persistent Personal Agents

👍 0

07/06 08:00

Persistent personal agents combine long-term memory with access to users' external environments, enabling personalized foreground assistance and proactive background execution. This integration also creates a new path to compromise: untrusted external content can be silently written into persistent

AgentGym2: Benchmarking Large Language Model Agents in De-Idealized Real-World Environments

👍 1

07/06 08:00

Language agents, i.e., LLM agents, progress rapidly and are increasingly deployed in production environments. This trend underscores the urgent need for rigorous and realistic evaluations. However, most existing benchmarks evaluate agents in simplified, idealized settings. They typically rely on pre

EdgeBench: Unveiling Scaling Laws of Learning from Real-World Environments

👍 10

07/06 08:00

Pretraining scaling laws reveal that model capability improves predictably with data and compute. But learning from real world environments after deployment remains far less understood. Analyzing roughly 38,000 hours of agent interaction with the environment across 134 real world tasks, we find, to

DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation

👍 1

07/06 08:00

Speculative decoding accelerates Large Language Model (LLM) inference by decoupling draft generation from target verification. While recent parallel drafters efficiently propose long token sequences in a single forward pass, they suffer from rapid acceptance decay due to a lack of inter-token depend

KVpop -- Key-Value Cache Compression with Predictive Online Pruning

👍 16

07/06 08:00

Key-value (KV) cache growth is a major bottleneck in autoregressive decoding, as memory and bandwidth scale linearly with context length. Existing KV eviction methods often rely on static heuristics or proxy scores, which poorly track future token utility and cause brittle eviction as relevance shif

Your Agent's Memories Are Not Its Own: Forged Reasoning Attacks on LLM Agent Memory and Defenses

👍 0

07/06 08:00

Persistent memory has enabled large language model (LLM) agents to store factual knowledge, prior decisions, reasoning histories, tool usage information, and context. While this has improved the agent's functionality and continuity across tasks, it has also introduced a new attack surface: the agent

InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization

👍 18

07/06 08:00

Unified models for robot manipulation aim to equip one policy with both the semantic priors of pretrained VLMs and the physical dynamics learned through future prediction. In practice, existing designs tend to erode the semantics of the pretrained backbone, suffer interference among heterogeneous ob

Qantara: Bridge-Flow Training for Multi-Paradigm JEPA Control

👍 0

07/06 08:00

Joint-Embedding Predictive Architectures (JEPAs) underpin a growing family of latent world models for control from raw pixels, but every existing JEPA world model commits at training time to a single inference paradigm: either trajectory optimisation in a learned dynamics model, or direct behaviour

Teaching LLMs a Low-Resource Language: Enhancing Code Completion in Pharo

👍 0

07/06 08:00

Large Language Models (LLMs) unlocked new possibilities in automated code writing, becoming the backbone of most code completion tools. While LLMs excel in mainstream languages, they often lack support for the so-called low-resource languages where training data is scarce. As a result, these languag

HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better

👍 0

07/06 08:00

We present HunyuanOCR-1.5, a lightweight end-to-end OCR-specialized vision-language model. HunyuanOCR unifies document parsing, text spotting, information extraction, text-image translation, and multi-image document understanding within a single end-to-end VLM. Building upon the lightweight architec

PAGE: Towards Practical Human-level Gaze Target Estimation

👍 0

07/06 08:00

Gaze target estimation, the task of predicting where a person is looking in a scene, is crucial to understanding human attention and intent. It is a challenging task that combines high-level understanding of global scene semantics and precise spatial reasoning using human appearance (e.g. pose, eye

URSA: Chemistry-Aware Benchmark for Utilitarian Retrosynthesis Assessment

👍 0

07/06 08:00

Synthesis planning aiming to find pathways of reactions for a target molecule is one of the most important and challenging tasks in drug discovery. Recent progress has produced both specialized deep-learning retrosynthesis systems and general-purpose large language models, but objective comparison r

Do All Visual Tokens Matter Equally? Object-Evidence Preserving Token Merging for Vision-Language Retrieval

👍 15

07/06 08:00

Multi-vector vision-language retrieval preserves fine-grained visual evidence through maximum-similarity late interaction, but dense image-side tokens make storage and scoring expensive. Existing token compression methods reduce this cost, yet they can remove or collapse object- and region-level evi

MTEB-PT: A Text Embedding Benchmark for Brazilian Portuguese

👍 0

07/06 08:00

Text embeddings for Portuguese have no dedicated benchmark: evaluation rests on translated corpora such as English MS MARCO or on thin multilingual coverage, with native tasks scattered and unconsolidated. We introduce MTEB-PT, a benchmark of 22 native Brazilian-Portuguese tasks across seven categor

Golden Hour Divide: Trauma Care Accessibility and Resource Vulnerability in Sri Lanka

👍 1

07/06 08:00

Timely intensive care dictates survival, yet emergency infrastructure remains unevenly distributed across Sri Lanka. While pre-hospital services have expanded, the transition to definitive care remains a critical bottleneck. This study evaluates national emergency resilience by quantifying the gap b

Wan-Streamer v0.2: Higher Resolution, Same Latency

👍 24

07/05 08:00

We present Wan-Streamer v0.2, a latency-preserving upgrade of the native-streaming, end-to-end audio-visual interaction model. v0.2 keeps the v0.1 modeling formulation, but raises the interactive output stream from 192x336 to 640x368 while preserving approximately 200 ms model-side signal-to-signal

Badge

07/08 05:13

AI agents collect peer reviews to generate proof of work

中文介绍 「Badge」是一款基于AI技术的平台，利用AI代理自动化收集同行评审，从而生成可信赖的工作证明。其核心功能在于简化传统工作验证流程，通过整合多方反馈，为用户提供客观、透明的专业能力和项目成果证明。

Scribble Network

07/08 05:13

The product that makes AI recommend your brand

中文介绍 「Scribble Network」是一款利用人工智能技术来推广品牌的产品。它旨在通过AI算法，优化品牌在网络中的可见度和推荐度，帮助企业和个人品牌更有效地触达目标受众，提升市场影响力。

LongCat-2.0

07/08 05:12

1.6T MoE trained entirely on AI ASICs

中文介绍 「LongCat-2.0」是一个大规模AI模型，采用MoE（专家混合）架构，其参数量达到1.6万亿。该模型的一大特点是完全基于AI专用集成电路（ASIC）进行训练，这表明其在训练效率和性能上可能具有独特优势，代表了AI硬件与模型协同发展的新方向。

Social Fetch

07/08 05:11

Social media scraper API for every major platform

中文介绍 「Social Fetch」提供一个社交媒体数据抓取API，旨在为用户提供从所有主流社交平台获取数据的能力。该API允许开发者和企业集成其服务，以便高效地收集、分析来自Facebook、Twitter、Instagram等平台的数据，支持市场研究和内容监控等应用。

Typeahead 2.0

07/08 05:08

Private AI autocomplete for every app on your Mac

中文介绍 Typeahead 2.0是一款为Mac用户设计的私有AI自动补全工具。它能为Mac上运行的任何应用程序提供智能文本预测功能，旨在提高用户输入效率，同时强调数据的本地处理和隐私保护。这款应用将AI技术融入日常操作，提升Mac的使用体验。

Sunrise

07/08 05:03

A real planner for Google Tasks

中文介绍 Sunrise是一款专为Google Tasks设计的“真实规划器”工具。它旨在增强Google Tasks的功能，提供更全面、更直观的任务管理和日程规划体验，帮助用户高效组织待办事项，提升个人和团队的工作效率。

Kadoink AI

07/08 05:02

Gather people instantly by AI txt, video, or ringing phones

中文介绍 「Kadoink AI」是一款利用人工智能技术即时召集人群的工具。它支持通过AI生成的文本消息、视频通话或直接拨打电话等多种方式，高效地联系和聚集目标群体。该产品旨在简化会议召集、活动组织或紧急通知等场景下的沟通流程。

Zoho Tables

07/08 05:02

Smarter way to manage work and data.

中文介绍 「Zoho Tables」是一款智能工作和数据管理平台。它提供了一个更高效、更直观的方式来组织、追踪和协作处理项目与数据。该产品旨在帮助用户提升工作效率，优化数据管理流程，适用于团队协作和个人生产力提升。

Cadence

07/08 05:02

Record once, send confidently

中文介绍 「Cadence」是一款旨在简化信息传递流程的工具。其核心理念是“录制一次，自信发送”，意味着用户可以高效地创建内容（例如录音或视频信息），并确保其能够准确、专业地传达给目标受众。该产品可能适用于销售、客服或内部沟通场景。

Katalyst

07/08 05:02

The AI agent that works your Salesforce Pipeline

中文介绍 「Katalyst」是一款专为Salesforce用户设计的AI代理。该产品旨在通过人工智能技术自动化和优化销售流程（Salesforce Pipeline），帮助销售团队更有效地管理潜在客户、推进交易，从而提升销售效率和业绩。

Claude Fable 5 Just Changed Design Forever

07/07 01:37

中文介绍 这则YouTube视频由Riley Brown发布，探讨了设计师为何应该转用“Claude Fable 5”。视频可能详细分析了“Claude Fable 5”这款工具或AI模型的功能、优势以及它如何提升设计工作效率，旨在说服设计师采纳此新平台。

Building a Real AI Jarvis (GPT Realtime 2 + Cursor)

07/01 09:40

Poweful Agents inside Slack

06/25 06:39

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

06/21 21:30

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

06/18 10:27

How to build high quality internal tools with Retool

06/18 05:21

SpaceX Just Bought Cursor for $60B. It’s About to Take OVER.

06/17 05:31

Why the Government Just Killed Claude Fable 5

06/13 11:50

This is Why They Banned Claude Fable 5

06/13 02:28

Hermes Agent NEW Super-App and DeepSeek v4 Catches Up To Opus 4.8?

06/07 04:10

几年前，大家会想到今天的 AI 那么离谱吗？

07/08 00:28

11 回复 · 程序员 节点

国内的 coding plan 套餐到底行不行

07/07 19:48

8 回复 · 程序员 节点

🔥 CNBC：美国公司开始更多采用中国 AI 模型

07/07 19:40

23 回复 · 程序员 节点

Codex 马上到期了，求靠谱充值方案

07/07 19:29

6 回复 · 程序员 节点

国行 iPhone air 在国内成功添加 giffgaff esim，顺便解锁了 apple news 等

07/07 17:37

10 回复 · Apple 节点

MAC 外接 32 寸显示器求推荐

07/07 14:52

7 回复 · Apple 节点

App Store 购买的软件需要登陆账号和购买账号保持一致了？

07/07 10:22

21 回复 · Apple 节点

mac 的外接显示器要完美必须 5K 吗？

07/06 15:11

67 回复 · Apple 节点

出个人闲置电脑

07/06 12:46

23 回复 · Apple 节点

[送永久激活码]RightX 增强你的 Mac Finder 右键

07/06 10:40

133 回复 · Apple 节点

All Cars Sold in the EU Now Require a Camera Aimed at Your Face

07/08 04:50

65 points · 65 comments

Show HN: Davit, a Apple Containers UI

07/08 02:44

Mostly vibe-coded Apple Containers front-end that I'd like to use myself. But if others want to use it, here's the source code.

AI Meets Cryptography 1: What AI Found in Cloudflare's Circl

07/08 02:36

45 points · 5 comments

Astro 7.0

07/08 02:30

133 points · 34 comments

Local, CPU-Friendly, High-Quality TTS (Text-to-Speech) with Kokoro

07/08 02:24

100 points · 20 comments

Show HN: Docx-CLI: agents read/edit Word docs using 1/2 the time and tokens

07/08 02:19

27 points · 10 comments

Notes on Software Quality

07/08 02:14

41 points · 24 comments

l: A new runtime for k and q

07/08 02:08

https://lv1.sh/blog/why-l/

China sentences official to death for taking $325M in bribes

07/08 00:38

209 points · 228 comments

Jim's TrueType QR Code Font

07/08 00:30

https://qr.jim.sh/

MacSurf 1.68 – NetSurf on OS 9 Released

07/08 00:24

57 points · 8 comments

Show HN: Rowboat – Open-source, local-first alternative to Claude Desktop

07/08 00:10

Claude’s desktop app is brilliant, but for our own daily work we kept wanting it to be less like a chat app and more like a full-fledged work app. Rowboat is our attempt at that, including the ability to build your own work surfaces inside Rowboat (more below).Our repo is https://github.co

30papers.com – Ilya's 30 essential ML papers, in a beginner friendly format

07/07 23:58

241 points · 39 comments

Why we built yet another Postgres connection pooler

07/07 23:36

87 points · 16 comments

Microsoft fire idTech team at Id software

07/07 23:33

419 points · 411 comments

Chat Control passed first round in EU Parliament

07/07 23:16

469 points · 203 comments

Automating AI Away

07/07 23:11

72 points · 37 comments

The revenge of the philosophy majors

07/07 22:41

109 points · 170 comments

Chat Control 1.0 and 2.0 Explained

07/07 22:23

281 points · 90 comments

Show HN: Halo – open-source, tamper-evident runtime evidence for AI agents

07/07 22:07

Hi HN, I'm Brian, I spent the last few years at Vanta (YC W18), helping startups and enterprises become compliant and I recently started exploring what that might look like in a post-agentic world.The problem Halo solves is: when a company buys an AI agent from a vendor and gives it access to t

98% isn't much

07/07 20:45

426 points · 274 comments

A better way to tie gym shorts (or any drawstring) [video]

07/07 20:45

403 points · 149 comments

StreetComplete: Fixing OpenStreetMap, one tiny quest at a time

07/07 20:38

600 points · 138 comments

9 Mothers (YC P26) Is Hiring in Austin, TX

07/07 20:00

1 points · 0 comments

Why skilled workers come to Germany and then leave again

07/07 18:42

114 points · 252 comments

GitHub Freno: cooperative, highly available throttler service

07/06 10:06

32 points · 0 comments

Fixing analog audio on the $2.58 HDMI-to-VGA adapter

07/05 13:36

49 points · 18 comments

Camera with transparent display launches for the equivalent of $29

07/04 04:54

14 points · 6 comments

Herdr: One terminal to rule them all

07/02 12:28

45 points · 29 comments

Computational Balloon Twisting: The Theory of Balloon Polyhedra [pdf]

07/02 07:26

26 points · 0 comments

v2.1.203

07/08 05:06

What's changed Added a warning when your login is about to expire, so you can re-authenticate before background sessions are interrupted Added a grey ⏸ badge to the footer when in manual permission mode, making the active mode always visible Added the session's additional working directories to MCP

中文介绍 Anthropic的Claude Code发布了v2.1.203版本更新。新版本主要进行了用户体验改进，包括添加了登录即将过期时的警告提示，以便用户能及时重新验证身份，避免后台会话中断。此外，在手动权限模式下，页脚会显示一个灰色暂停徽章，使用户能清晰看到当前模式状态。

v2.1.202

07/07 06:51

What's changed Added a "Dynamic workflow size" setting in /config for controlling how large Claude generally makes dynamic workflows (small/medium/large agent counts) — an advisory guideline, not an enforced cap Added workflow.run_id and workflow.name OpenTelemetry attributes to telemetry emitted by

中文介绍 Anthropic的Claude Code发布了v2.1.202版本更新。此次更新新增了在/config中配置「动态工作流大小」的选项，允许用户选择Claude生成工作流的代理数量规模（小型、中型或大型），这作为指导性设置。此外，新版本还为workflow.run_id和workflow.name添加了OpenTelemetry属性，以增强可观测性。

v2.1.201

07/04 07:50

What's changed Claude Sonnet 5 sessions no longer use the mid-conversation system role for harness reminders

中文介绍 克劳德（Claude）代码库发布了 v2.1.201 版本。此版本更新的一项主要变化是，Claude Sonnet 5 的会话在进行引导提醒时，将不再使用对话中的系统角色。这反映了对模型内部交互机制的调整和优化，旨在提升模型性能和用户体验。

v2.1.200

07/04 00:52

What's changed Changed AskUserQuestion dialogs to no longer auto-continue by default; opt into an idle timeout via /config Changed the "default" permission mode to "Manual" across the CLI, --help, VS Code, and JetBrains; --permission-mode manual and "defaultMode": "manual" are accepted alongside def

中文介绍 Anthropic的Claude Code发布v2.1.200版本。此版本更新了“AskUserQuestion”对话框，默认不再自动继续，用户可通过“/config”设置空闲超时。此外，CLI、VS Code和JetBrains等环境的默认权限模式均已更改为“手动”，以提高安全性控制。

v2.1.199

07/03 07:35

What's changed Stacked slash-skill invocations like /skill-a /skill-b do XYZ now load all leading skills (up to 5), not just the first Fixed SSL certificate errors (TLS-inspecting proxies, missing NODE_EXTRA_CA_CERTS, expired certs) burning retries before showing actionable guidance — they now fail

中文介绍 Anthropic的Claude Code发布v2.1.199版本。此更新改进了堆叠斜杠技能调用功能，现在最多可加载前5个技能，而非仅第一个。同时，修复了与SSL证书相关的错误，解决了因TLS检查代理、缺少NODE_EXTRA_CA_CERTS或证书过期导致的重试问题。

v2.1.198

07/03 06:55

What's changed Subagents now run in the background by default, so Claude keeps working while they run and is notified when they finish (previously a gradual rollout) Claude in Chrome is now generally available Added background agent notifications in claude agents — sessions that need input or finish

中文介绍 Anthropic的Claude v2.1.198版本更新。Claude现已在Chrome浏览器上普遍可用，并为Claude代理新增后台通知功能，在会话需输入或结束时触发。同时，增加了"/dataviz"技能，提供图表和仪表盘设计指导。

v2.1.197

07/01 01:56

What's changed Introducing Claude Sonnet 5: now the default model in Claude Code, with a native 1M-token context window and promotional pricing of $2/$10 per Mtok through August 31. Update to version 2.1.197 for access. https://www.anthropic.com/news/claude-sonnet-5

v2.1.196

06/30 07:27

What's changed Added support for organization default models — admins set it in the org console; it shows as "Org default" (or "Role default") in /model when you haven't picked one yourself Added readable default names for sessions at start, making them easier to identify and message Added clickable

v2.1.195

06/27 05:29

What's changed Added CLAUDE_CODE_DISABLE_MOUSE_CLICKS to disable mouse click/drag/hover in fullscreen mode while keeping wheel scroll Fixed hook matchers with hyphenated identifiers (e.g. code-reviewer, mcp__brave-search) accidentally substring-matching — they now exact-match. Use mcp__brave-search_

v2.1.193

06/26 05:45

What's changed Added autoMode.classifyAllShell setting to route all Bash/PowerShell commands through the auto-mode classifier instead of only arbitrary-code-execution patterns Added auto-mode denial reasons to the transcript, the denial toast, and /permissions recent denials Added claude_code.assist