1
Indian AI coding startup Emergent becomes a unicorn with $130M Series C
印度AI编程新星成立仅一年多，C轮融资1.3亿美元估值飙至15亿美元，成为独角兽。
The startup has reached a $120 million annualized revenue run rate and more than 200,000 paying customers.
2
Agent runtime security: Foundry, GitHub, Mastra updates
Azure Agent运行时安全更新：Foundry、GitHub与Mastra工具新特性，带你了解Procedural memory和Toolboxes公开预览与Teams GA计划。
This week drew a sharp line between building agents and running them safely in production. Two significant supply chain and trust-boundary failures la…
3
Authoring for Living Worlds: Tool-Constrained LLM Agents for Executable Multi-Actor Scenarios
AI代理自动生成3D游戏可执行的叙事场景，让非专业人士也能创作动态多角色视频。
arXiv:2604.10383v2 Announce Type: replace Abstract: We use LLM agents to author executable specifications for a living world: formal Graphs of Events …
4
Catalyst-Agent: Autonomous heterogeneous catalyst screening with an LLM Agent
用LLM代理自主筛选电催化剂，MLIPs和图神经网络助力，加速新材料发现。
arXiv:2603.01311v3 Announce Type: replace Abstract: The discovery of catalysts for electrochemical applications such as the oxygen reduction reaction …
5
Show HN: Backscroll – search your ChatGPT/Claude/Gemini history
再也不用翻聊天记录，一键搜索ChatGPT、Claude、Gemini全部历史对话
Article URL: https://backscroll.xyz Comments URL: https://news.ycombinator.com/item?id=48920142 Points: 2 # Comments: 0
6
出版商与作者集体起诉谷歌，指控其盗用版权内容训练 Gemini AI
出版商和作家组团起诉谷歌，指控其未经授权用版权作品训练Gemini AI，类似案件曾引发天价赔偿，版权与AI的博弈再升级。
IT之家 7 月 15 日消息，一群出版商和作家已对谷歌提起集体诉讼，指控这家科技巨头未经授权使用他们的版权作品训练其人工智能平台 Gemini。 原告方包括 Hachette Livre、Cengage Group、Elsevier、作家 Scott Turow 以及 S.C.R.I.B.E . …
7
Show HN: DevQuest – live GitHub profile cards and a daily repository battle
实时生成GitHub个人资料卡片，还有每日仓库对战模式，让你的README动起来！
I built DevQuest, a tool for generating live GitHub profile cards for READMEs. I also added a Daily Battle mode where developers can submit a GitHub r…
8
The founder of Hinge raised $18M to build a new AI dating service, Overtone
Hinge创始人获1800万美元融资，打造AI驱动的语音约会新服务Overtone，不是传统app！
Overtone describes itself as "a voice- and audio-forward service, enabled by AI, that provides highly curated introductions."
9
Show HN: Agent's Design – Claude/Codex copy-paste templates that kill AI-slop UI
专为Claude/Codex设计的复制粘贴模板，轻松告别AI生成的低质UI。
Article URL: https://agents-design.com Comments URL: https://news.ycombinator.com/item?id=48920581 Points: 3 # Comments: 0
10
联想海外推出摩托罗拉 Edge 70 Max 手机：骁龙 8 Gen 5 处理器、7100mAh 电池
骁龙8 Gen 5+7100mAh电池，IP69认证，摩托罗拉海外新机亮点十足
IT之家 7 月 15 日消息，联想现已在海外推出了摩托罗拉 Edge 70 Max 手机，该机搭载高通骁龙 8 Gen 5 处理器，匹配 7100mAh 电池， 起售价为 54999 印度卢比（IT之家注：现汇率约合 3901 元人民币） 。 该机可选三种配色，拥有 IP69 和 MIL-STD-…
11
ChatGPT Subagents 2
OpenAI推出Subagents 2，却被吐槽仅是移除透明度和加密输出，毫无新意。
openai mfs were like yall liked subagents? introducing: subagents 2!!! and then its just the same thing with fully removed transparency and all output…
12
Hackers quickly prove that Neo Geo Doom ports are not "impossible"
黑科技巧突破硬件极限，经典游戏《毁灭战士》成功移植Neo Geo，证明技术没有"不可能
Clever coding and graphical compromises get a classic game on more classic hardware.
13
Agentic Context Learning with Self-Discovered Specification
让AI自动发现任务规范，在上下文中动态学习，实现更强自主推理能力。
arXiv:2607.09794v1 Announce Type: new Abstract: Context learning is an emerging inference-time task where LLMs must learn and apply novel, task-specif…
14
15
How to Build Your First AI Agent With Claude Code in 30 Minutes
30分钟用Claude Code搭建首个AI代理，学会用目标驱动代替步骤指令，效率翻倍。
Most "build your first agent" tutorials start with a framework and four new vocabulary words before you've written anything that does something. This …
16
AI Wrote Your Code. Did It Ship a Vulnerability Too?
Veracode报告揭示AI生成代码安全瓶颈，新模型GPT-5/Claude Sonnet 4.5通过率仅50-55%，模型规模增大并未提升安全性。
TL;DR Nearly half of AI-generated code fails basic security tests — and that number hasn't budged as models got bigger. Java is the riskiest language,…
17
MemDecay: Region-Aware KV Cache Eviction for Efficient LLM Agent Inference
新方法MemDecay基于区域感知的KV缓存淘汰，大幅提升LLM Agent推理效率，降低内存占用。
arXiv:2607.10582v1 Announce Type: cross Abstract: Large language model (LLM) agents accumulate heterogeneous context, including system instructions, p…
18
Automated Textbook Auditing with Multi-Agent LLM Systems
多智能体大模型系统自动审核教科书，AI赋能教育内容质量把控新范式
arXiv:2607.11276v1 Announce Type: cross Abstract: Ensuring the quality of educational materials requires more than standard proofreading: textbooks mu…
19
Show HN: Liveshortly - stream and pair prompt with ai agents
实时流式配对AI代理，轻松管理多Agent协作，提升效率新工具
We built a tool for livestream your agents session and it also allows collaborative pair prompting with team mates in single claude or codex sessions.…
20
Agent Hacks Agent: Autoresearch for Production-Agent Red-Teaming
让AI Agent自动互相攻击，发现生产环境漏洞，这篇论文提出了全新的自动化红队测试框架。
arXiv:2607.11698v1 Announce Type: cross Abstract: Production LLM agents such as Claude Code and Codex operate over untrusted content, files, commands,…