Tag
The author shares a retrospective of using their own subagent-team skill to lead a Claude/Codex/etc. AI team through a major monorepo overhaul: cross-vendor model reviews offer the best ROI, rework volume depends on task risk rather than model tier, and they summarize a division of labor where low-tier models act as coordinators, strong models handle risky implementation, and top-tier models handle reviews.
An educational walkthrough of what happens between a user's message and a chatbot's reply, covering tokenization, context windows, conversation memory, chunking/embeddings, RAG retrieval, system prompts, function calling, and safety evaluation.
OccupySaaS 是一场由被裁员工程师发起的抗议运动与开源项目，主张用 AI 从公开文档以 clean-room 方式重建专有 SaaS 的替代品，首批发布 ZeroAuthy（Auth0 兼容身份服务器）和 SalesCompel（Salesforce 替代 CRM），均采用 Apache-2.0 许可。
A developer announced HarbourLine, a cozy simulation game where players grow a fishing wharf into an automated megaport, built entirely with Claude Code / Opus 5.5 including its trailer.
Kent Beck argues that trust, not code generation speed, is the binding constraint on software delivery in the AI era, and that XP practices like testing and pairing manufacture the confidence needed to release; he also promotes his upcoming paid class on trust-optimized AI-assisted development.
作者感叹 AI 领域变化过快，积累的笔记和书签很快过时，且在 agentic coding、提示词实践、MCP/技能等工具选择上缺乏共识，每次新模型发布都会让旧经验作废。
A commentator shares thoughts on a VC-funded experiment pushing state-of-the-art AI to ship code directly to production in a game environment where failures are tolerable.
Beijing has established the National Data Group Co, a state-owned enterprise headquartered in Xiongan New Area, to aggregate, manage, and commercialise national data assets in support of China's AI development ambitions, centralising data governance to reduce fragmentation and bolster data sovereignty.
A poster shares feelings of loneliness and anxiety about the accelerating pace of AI development, the risks of rogue agents, and societal impact, asking others if they share similar concerns.
A community discussion asking whether vibe coding with AI has actually produced profitable, real-user products or mostly just demos, and whether zero-experience builders can genuinely launch successful apps and become independent developers.
A Big Tech engineer compared building the same open-source library with DeepSeek V4-flash ($19.86) versus Opus 5.5 ($346.70), finding an 18x cost difference with acceptable results, arguing the industry should reconsider model rightsizing.
作者认为AI驱动的软件开发中，评测系统（evals）正在取代传统自动化测试，成为提升开发效率和系统质量的关键，并提到Claude Code团队和Google的harness2等大厂都在加大投入。
The author shares a human-in-the-loop continuous development workflow based on Claude Code as the main agent orchestrating Codex CLI, Subagents, and GitHub Issues/PRs, covering repo initialization conventions, task decomposition, and review/merge processes, along with reusable prompt templates.
作者分享购买 ESP32 桌面机器人并用 AI 辅助开发的体验，感慨如今只需连接 Type-C 就能让 AI 完成代码迭代，做各种有趣的智能小装置。
trq212 and fnthawar discuss how the key skill when working with coding agents is 'being in the details'—like running a company, you can't know everything, but obsessively caring about the right details (code, architecture, design, UX) is what makes something great.
Google dominates by the key metric for sustaining long-term AI development, such as compute capacity and infrastructure scale, reinforcing its leading position in the AI race.
John Rush 分享了他用 Claude Code 在 5 分钟内生成一个分步模拟向导表格的经历，并认为这种 vibe coding 方式可应用于建房、修车、手术等任何实体技能场景。
A developer (@huangtongxueh) reflects on gaining over 1,000 followers and explains their "learn in public" writing process, covering how they use AI tools like Codex, Claude, and Agents in daily development to distill reusable workflows and articles.
Developers reportedly used large language models ("vibe coding") to generate an NVIDIA graphics card driver that runs on modern macOS, reviving GPU support Apple abandoned after High Sierra. The feat is seen as evidence that AI could break down long-held closed-source moats built by companies like Apple and NVIDIA.
作者分析为什么编码智能体（如 Claude Code、OpenCode、Codex）表现迟钝：模型本身能力很强，但智能体作为连接模型与代码库的"身体"在任务管理、并行执行和委派方面严重滞后，成为 AI 辅助开发的瓶颈。