AI 教程与实战

I sat in on Margaret-Anne Storey's DORA community session last week, and she put a name on the thing...

GitHub Repository: https://github.com/13manmayarai-hash/omnicache-proxy (MIT Licensed) What...

Before asking an AI coding agent to “fix this bug,” there is one question worth asking first: Does...

One of our product levers is capped at three executions per ISO week. In week 2026-W35 the ledger...

Previously: 9 Bugs That All Looked Like a Working System · I Built an AI That Rewrites Its Own...

LangChain 发布指南，介绍 Schneider Electric、Vodafone 和 monday.com 在欧洲与中东规模化落地生产级 AI 的做法。内容涵盖建立共享 Agent 平台与 LLMOps 实践，以及设计具备更强可观测性、评估和控制能力的多智能体架构。

Three gotchas installing a plugin into someone else's agent This one is scoped tight:...

The discriminator between a membrane and a logger The first version of this test passed...

Meta 首席 AI 官 Alexandr Wang 转发用户 @jaystar524 的实测反馈，称 muse spark 1.3 表现可靠。该用户用多个 sub-agents 并行完成任务，耗时约一小时。

Alexandr Wang 发帖号召大家试用 Muse Spark 1.3，并引用用户 @DeryaTR_ 的体验：用该模型快速构建了一个海盗平台跳跃游戏，一次生成即无缺陷运行，作者称其正在继续美化美术、精灵图并添加新关卡，认为这是一个快速、便宜、好用的编码模型。

LangChain 博客文章介绍 Nevermined 如何赋能 LangChain 智能体购买和出售服务，正文无更多细节。

Hugging Face 发布完整教程，使用 TRL 库以 GRPO 微调 LiquidAI/LFM2.5-350M，约 500 样本、100 训练步即可在免费级 Colab 或 Kaggle GPU 上完成，训练后模型在 IFStruct 基准上从 22.6% 提升至 29.7%。

研究者提出扩散增强LLM（diffusion-augmented LLMs），通过轻量扩散权重并行生成多个token，在不引入草稿模型或牺牲AR模型质量的前提下加速生成。配套的Ψ-Spec采样器实现无损加速，Uno模型吞吐量优于主流投机解码方法，较基础AR模型最高提速3倍。8B Uno在智能体工具使用、编程和长上下文推理基准上超越26B DiffusionGemma及Mercury 2。

EVOHARNESSBENCH 是一个新基准，用于评测智能体在工具、技能和智能体三个维度的受控框架演进下的表现。该基准包含 17 个多阶段流、802 个任务、520 个工具、42 项技能和 62 个智能体。结果显示，框架扩展本身会降低此前已解决任务的性能，产生框架引发的遗忘，且保留旧能力与适应新能力可能相互冲突。

伦斯勒理工学院研究团队提出"循环状态写回"概念，揭示低精度推理中状态存储规则对循环网络时序行为的关键影响。在荧光寿命成像的GRU编码器-解码器中，将连续状态传播替换为确定性4-bit状态存储，使τ1和τ2估计误差分别增加约70倍和300倍；误差反馈、残差记忆和方向记忆可在不重训练的情况下恢复精度。该现象在独立训练的LSTM中同样复现，表明状态存储接口是量化循环推理的核心设计考量。

Lint is green. The bot left forty comments about naming. Somebody typed LGTM. Then billing...

本期早报围绕能力、工作过程与最终成本展开，精讲三篇：Google 发布 Gemini 3.8 Flash 系列含通用 Flash 与 Flash Cyber，内部漏洞发现成功率超 70%。

The /journal skill from my breadcrumbs post is now a Claude Code plugin anyone can install. One glanceable callout per topic, colored by season with an icon for the weekday, three Notion calls per run. What it writes, how to set it up, and…

Before GPUs. Before neural networks. Before transistors were even a rumor. There was a hungry cat in...

Telegram exports two different shapes, and neither is the hard part. The hard part is packing messages under a 500k-word source cap and making the second run idempotent.

Every developer I know now runs an AI coding agent in something like auto-accept mode. Claude Code,...

Your coding agent finishes a task and tells you the tests passed. You accept the report. Later, when...

Hello, I'm Rijul, and I'm building LiveReview — a blast-radius aware AI code review built for your...

A review protocol that challenges the scheme first, then the code, then stops. Full prompt included.

Claude 官方频道发布视频，介绍 Claude Code 团队自己如何使用 Claude Code。材料仅含标题，无正文文本，具体内容细节以视频为准。

I build and sell Reestri, a small catalog of paid Apify actors that answer one kind of question: who...

GitHub 工程师 Erik Kristensen 分享了 Copilot 降本的四项改动：选择性压缩工具输出、移除 view 工具行号前缀（线下推理成本降约 5%，线上用户日均推理成本降约 3%）、压缩 task-tool 提示词（每轮省约 1300 token，每活跃小时归一化成本降 2.9%）、后台任务完成后直接交付结果（AI Credits 用量降约 2.3%）。

Most "AI-powered" side projects follow the same arc: a slick demo, a hardcoded API key, one shared...

How I added a deterministic trace contract to an AI Terraform reviewer so rejected infrastructure...

A portfolio website should make it easy for a potential client to answer three questions: What do you...

Quick answer: the best academic search API for AI agents in 2026 is Valyu, because it is the only one...

My hot take is: Telling people that they have to build endlessly is very, very, very bad advice. If...

Google AI 团队发布教程，讲解如何为 LLM-as-a-Judge 评测编写可靠的布尔式评分标准，指出模糊提示会导致评估不一致和浪费 token。文中给出四条经验：问题保持原子化且互不重叠、只让评判模型评估客观事实（可用 RFC 2119 术语如 MUST 表述）、只评 prompt 中明确要求的内容、用专家标注的 golden set 校准评判模型直至与人类评分一致。

If your AI application suddenly becomes 300–500ms slower after adding an AI gateway, the first...

Useful AI agents keep forcing a bad choice: lock the environment down until it is useless, or hand the agent your real machine. ShrekOS is my attempt to move that trust decision below the agent, into the operating system.

Anthropic removed over 80% of Claude Code's system prompt for Claude Opus 5 and Claude Fable 5, and...

百度智能云具身智能前沿讲堂上，清华大学AIR助理教授赵昊与百度智能云AI计算首席科学家王雁鹏等对谈世界模型规模化问题。赵昊提出以3D/4D Token为核心的世界模型路线，认为世界模型不等于视频生成，需表达空间结构、光学属性和速度、加速度等动态信息；其OmniNWM通过动作编码与相机参数解耦，实现200帧稳定生成，并与精锋医疗、极智嘉分别合作手术预演和仓库机器人场景。

Your coding agent is good at reading code. Point Claude Code or Cursor at a repo and it...

Since 2026 every Shopify storefront answers an agent-commerce endpoint at POST /api/ucp/mcp,...