1
Show HN: Live AI music sequencing agent
在网页上与AI实时共创音乐，所有听众同步体验同一首AI编排的旋律。
Pretzel is an experiment in live AI agents for music. It's a web synchronized music sequencer that is controlled by an AI agent. Everyone chats with t…
2
Skill Retrieval Augmentation for Agentic AI
让LLM智能体学会检索外部技能，摆脱显式枚举限制，提升复杂任务处理能力
arXiv:2604.24594v2 Announce Type: replace-cross Abstract: As large language models (LLMs) evolve into agentic problem solvers, they increasingly rely …
3
Parallel Context Compaction for Long-Horizon LLM Agent Serving
针对长时LLM Agent的上下文溢出问题，提出并行压缩方法，减少数十秒推理阻塞。
arXiv:2605.23296v1 Announce Type: new Abstract: Long-horizon LLM agents accumulate growing conversation histories that eventually exceed the model's c…
4
How to sandbox AI coding agents without crippling them
AI编码代理沙箱怎么做才不削弱能力？看这篇避开Docker常见坑，给出真正可行的安全边界方案
The Problem: Your AI Agent Has Root A few months back I was helping a team set up a self-hosted AI coding agent. Standard setup — an LLM with tool acc…
5
Multi-Repo Microservice Changes Are a Coordination Problem. I Solved It With AI Agent Teams.
用AI代理团队解决多仓库微服务变更的协调难题，实战案例拆解。
A walkthrough of RepoOrch — an open-source Claude Code plugin that turns a multi-repo workspace into a deliberating team of AI specialists, with peer-…
6
How does an AI agent pick from 686 skills in a second?
实测AI代理如何在一秒内从686个技能中精准选中最优解，准确率68%的秘密武器是语义路由+向量检索。
I ran an empirical test on the "skills as semantic router" pattern for Claude Code agents. I indexed 686 randomly sampled skills from a 4,556-skill co…
7
Constraint Decay: The Fragility of LLM Agents in Back End Code Generation
arXiv新研究揭示LLM代理在后端代码生成中面临"约束衰减"问题，系统脆弱性令人警醒。
Article URL: https://arxiv.org/abs/2605.06445 Comments URL: https://news.ycombinator.com/item?id=48256912 Points: 4 # Comments: 0
8
Show HN: I built a RAG and knowledge graph agent that runs locally
本地运行的RAG与知识图谱AI代理，一键解决隐私与安全痛点，无需依赖云端模型配置。
Claw-Coder is an AI agent that runs locally on your laptop and has access to powerful tools instead of configuring claude or codex to use a local mode…
9
Computer-Use-Linux
Linux桌面用户福音：开源工具让AI代理直接操控你的计算机，无需等待官方支持
Hi, Since apps like Codex and Claude don't provide support for computer use for Linux desktop users, iv'e built my own own one. https://github.com/age…
10
OpenAI named a Leader in enterprise coding agents by Gartner
OpenAI被Gartner评为企业AI编码代理领导者，Codex展现强大创新与规模化部署能力。
OpenAI is named a leader in the 2026 Gartner Magic Quadrant for Enterprise AI Coding Agents, with Codex recognized for innovation and enterprise-scale…
11
Show HN: I built a powerful RAG and knowledge graph agent that runs locally
本地运行、保护隐私的RAG+知识图谱智能代理，无需配置云模型即可使用强大工具。
Claw-Coder is an AI agent that runs locally on your laptop and has access to powerful tools instead of configuring claude or codex to use a local mode…
12
Spotify’s AI bet: more of everything, less of what you want
Spotify押注AI代理功能，却可能陷入“越多越不精”的个性化困境，产品体验的微妙平衡值得关注。
Spotify has released a bunch of AI-powered tools that nudge users to create more content. It can be a bit much.
13
Autonomous LLM Agents & CTFs: A Second Look
再看自主LLM智能体在CTF挑战中的表现，更新发现与能力边界。
arXiv:2605.21497v1 Announce Type: cross Abstract: Large Language Model (LLM) agents are increasingly proposed to automate offensive security tasks, wi…
14
Herdr: A tmux-like terminal multiplexer for AI coding agents
专为AI编码代理设计的终端复用器，如tmux般高效管理多个AI会话。
Article URL: https://github.com/ogulcancelik/herdr Comments URL: https://news.ycombinator.com/item?id=48247248 Points: 3 # Comments: 0
15
微软 35 年老将优素福 · 迈赫迪宣布离职，最后一年将协助 Windows 适应 Agent 时代
微软35年资深高管宣布离职，最后一年专注推动Windows接入Agent时代，AI战略转型信号明确。
IT之家 5 月 22 日消息，据《商业内幕》今日报道，在微软工作 35 年的资深高管优素福 · 迈赫迪将在下一财年结束后离开公司。2023 年至今，迈赫迪担任微软商业业务首席营销官。 迈赫迪在备忘录中写道：“在微软度过 35 段非凡岁月后，我认为现在是开始规划下一段人生旅程的时候了。在这 35 年…
16
AgentEscapeBench: Evaluating Out-of-Domain Tool-Grounded Reasoning in LLM Agents
魔幻庄园式评测，专测LLM代理能否在陌生工具场景下自主推理与修正操作。
arXiv:2605.07926v2 Announce Type: replace Abstract: As LLM-based agents increasingly rely on external tools, it is important to evaluate their ability…
17
ImProver: Agent-Based Automated Proof Optimization
基于智能体的自动证明优化框架，提升数学证明的简洁性与可读性，ICLR 2025 顶会论文。
arXiv:2410.04753v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have been used to generate formal proofs of mathematical theore…
18
Ask HN: Is the next big thing locally running coding agents?
当云端API成本飙升，本地运行开源模型完成编程任务或将成主流趋势。
There's extreme price escalation on part of Anthropic, with token spend now approaching levels that have made many-an-enterprise scratch their heads. …
19
Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents
系统级诊断LLM代理执行失败，从单条追踪到全语料，规模化发现隐藏错误模式。
arXiv:2605.21347v1 Announce Type: cross Abstract: Diagnosing failures in LLM agents remains largely manual. Practitioners inspect a small subset of ex…
20
FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents
用语言代理实现大模型自主微调的创新框架，省去人工干预，让微调过程自动化
arXiv:2603.01712v2 Announce Type: replace-cross Abstract: Fine-tuning large language models for vertical domains remains labor-intensive, requiring pr…