1
让Skill“有图可依”：openJiuwen首发多模态Skill范式Skill-Omni
openJiuwen首发多模态Skill范式Skill-Omni，让Agent从“读得懂”升级为“看得见”，将视觉知识沉淀为可复用的技能。
从文字说明书到多模态经验库
2
SkillFab: An Agent-Native Skill Production Platform
面向AI代理的原生技能生产平台SkillFab，让技能制作更智能高效。
arXiv:2607.03780v1 Announce Type: cross Abstract: SkillFab is an agent-native platform for turning missing capabilities into reviewed, reusable Agent …
3
OpenSquilla发布0.5.0 Preview：多模型集成登顶DRACO双榜，对比名单中出现最新旗舰Fable 5
集成多模型并智能路由，自动选模型省钱又高效，登顶DRACO双榜，平均成本仅$0.12
少烧钱、真交付
4
Show HN: Skillmaxxing – make every agent self-evolving
让AI agent自动复盘工作，把经验沉淀为可复用的"技能"，无需手动触发，几天内自动生成4个合理技能，越用越强。
we built an agent plugin that lets an agent reflects after real work saves the reusable part as a "skill", or improves one it already has. No command,…
5
Cisco AI Defense Skill Scanner
一款由思科开源的AI技能安全扫描器，5分钟快速检测提示注入与数据泄露风险。
Article URL: https://github.com/cisco-ai-defense/skill-scanner Comments URL: https://news.ycombinator.com/item?id=48656076 Points: 2 # Comments: 0
6
What Is SKILL.md? A Practical Guide to AI Agent Skills
从单一公司功能到多平台标准，一文说清SKILL.md如何成为AI编码代理的核心技能定义文件。
If you work with AI coding agents, you have probably seen SKILL.md mentioned and wondered what it actually is. In a matter of months it went from one …
7
微信支付发布 AI 接入工具箱 2.0，Token 消耗少一半
微信支付AI接入工具箱2.0，Token消耗减半，全线覆盖支付与券场景，时序图转Mermaid格式，高效安全易集成。
IT之家 6 月 15 日消息，今年 4 月， 微信支付官宣发布一整套面向 AI 的支付接入 Skill ，涵盖 Skill 技能包、AI 友好文档、AI 友好 API 三大工具。据微信公开课消息，微信支付 AI 接入工具箱迎来 2.0 版本，围绕“安全、高效、原生流畅”的方向，重点更新了五大内容。…
8
SkillAudit: Ground-Truth-Free Skill Evolution via Paired Trajectory Auditing
无需真实标签，通过配对轨迹审计实现技能演化，为智能体能力进化提供新范式
arXiv:2606.14239v1 Announce Type: new Abstract: Agent skills are structured procedural packages that guide frozen LLM agents in specialized workflows.…
9
SkillChain: Closing the Loop on Skill Evolution for Image-Based E-Commerce AI Assistants
电商AI助手技能自主进化新框架，关闭闭环实现动态成长，图像电商场景再升级。
arXiv:2606.12984v1 Announce Type: new Abstract: Image-based AI assistants are now deployed at production scale on e-commerce platforms, where a single…
10
11
AI Builder Notes - Week of June 8, 2026
Vercel新API实现60万技能搜索与OIDC认证，代理部署误区引深思。
AI-assisted notes from my liked-tweets feed, organized around agent loops, cloud agent infrastructure, skill security, memory, and runtime context. Tr…
12
Skill-RM: Unifying Heterogeneous Evaluation Criteria via Agent Skill
这篇论文提出Skill-RM框架，通过智能体技能统一异构评估标准，推动AI评估体系革新。
arXiv:2606.03980v1 Announce Type: cross Abstract: Reward models (RMs) provide critical feedback signals for LLM post-training, notably in reinforced f…
13
14
When Safe Skills Collide: Measuring Compositional Risk in Agent Skill Ecosystems
社区贡献的AI技能单独安全，组合却可能引发风险？本文提出SkillReact量化测评这一新安全问题
arXiv:2606.00448v1 Announce Type: cross Abstract: LLM agents increasingly rely on community-contributed skills that expand an agent's operational capa…
15
SkillBrew: Multi-Objective Curation of Skill Banks for LLM Agents
让LLM Agents拥有更强大的技能库，多目标策划方法SkillBrew带来更高效的技能筛选与组合方案。
arXiv:2605.29440v1 Announce Type: cross Abstract: Retrieval-augmented LLM agents increasingly rely on curated skill banks: collections of reusable tex…
16
Claude Code as a Daily Driver: Claude.md, Skills, Subagents, Plugins, and MCPs
Claude Code进阶攻略：掌握CLAUDE.md、Skills和Subagents，让AI编码更得心应手。
Article URL: https://arps18.github.io/posts/claude-code-mastery/ Comments URL: https://news.ycombinator.com/item?id=48289950 Points: 296 # Comments: 2…
17
SkillGrad: Optimizing Agent Skills Like Gradient Descent
像梯度下降优化参数一样优化AI代理技能，全新SkillGrad方法带来更高效的自适应学习范式。
arXiv:2605.27760v1 Announce Type: new Abstract: Agent skills provide a lightweight way to adapt LLM agents to specialized domains by storing reusable …
18
美团推出“跑腿 Skill”，可对接各大 AI 助手实现“一句话点单”
对接各大AI助手，用自然语言一键下单跑腿服务，自动识别场景并完成订单，省去多步操作。
IT之家 5 月 26 日消息，美团发文宣布推出“跑腿 Skill”，相应 Skill 主要用于对接各大 AI 助手，接入后用户只需利用自然语言输入命令，即可让 AI 助手调用美团跑腿完成下单，系统自动完成场景识别、地址匹配、价格预估与订单提交，将原本多步操作压缩为一步。 据介绍，该 Skill 采…
19
OpenSkillEval: Automatically Auditing the Open Skill Ecosystem for LLM Agents
新论文提出自动审计框架，评估LLM智能体的开放技能生态系统，助力安全与能力验证。
arXiv:2605.23657v1 Announce Type: new Abstract: Skills, i.e., structured workflow instructions distilled for large language models (LLMs), are becomin…
20
Skill Weaving: Efficient LLM Improvement via Modular Skillpacks
新框架SkillWeave通过模块化技能包让大模型高效专业化，成本低、兼容多领域
arXiv:2605.22205v1 Announce Type: new Abstract: Large language models increasingly require specialization across diverse domains, yet existing approac…