Published every day
Daily updates
Three threads, published automatically every day: the briefing covers today, deep dives explain one topic, the course builds skills.
AI Briefing
dailyToday's AI world in five minutes: releases, agent engineering, research and industry.
AI Deep Dives
dailyOne technical topic a day, written up as a thorough deep-dive document.
- 2026/10/23Agent Skills 移交 AAIF:一个跨厂商技能包标准的归属、规范与风险 | Agent Skills Moves to AAIF: Ownership, Spec and Risks of a Cross-Vendor Skill FormatNew
- 2026/10/22Patch the Planet:AI 发现的漏洞怎么协调披露 | Patch the Planet: Coordinated Disclosure for AI-Found VulnerabilitiesNew
- 2026/10/21Anthropic 反蒸馏防御:推理轨迹怎么被保护 | How Anthropic Defends Claude's Reasoning Against DistillationNew
- 2026/10/20怎么验证 Agent 做的代码迁移“行为没变”:OpenAI MultiLangBench 与 COBOL 迁移评测 | How to Check an Agent's Code Migration Kept Behavior: OpenAI's MultiLangBench and COBOL Migration EvalsNew
- 2026/10/19银行为什么拦外部 AI Agent:凭证代登录 vs 受限接入 | Why Banks Block Outside AI Agents: Credential Login vs Scoped Access
Claude Code Lessons
dailyOne lesson a day, grounded in the official docs and changelog: one technique you can use today.