Skip to content

#Anthropic

0 items today
9/29Tue
  3. JavaGuide71

    NVIDIA 开源 Agent 沙箱项目 OpenShell

    NVIDIA 开源了 Agent 运行环境 OpenShell，用策略规定 Agent 能碰哪些文件、访问哪些服务，并在实际操作发生时执行这些规则。请求先经过沙箱识别发起程序，再由沙箱外的 Supervisor 核对策略后才建立外部连接；凭据由 Provider 管理，Agent 环境里拿到的是占位令牌，Proxy 转发前才替换成真实凭据。

    Awaiting translation

  5. Habr · Claude Code82

    A Product Designer Went Solo with Claude Code for a Month: What I Built Around the AI to Keep the Project from Falling Apart

    A product designer with six years of SaaS experience used Claude Code to single-handedly build Котомка, a life-planning app. Nearly all the code was written by AI; his job was to define requirements, review the results, and make decisions.

    Why it matters: Using a real repository, the author documented the pitfalls he hit while building a product on Claude Code alone, plus the rules, hooks, and testing guardrails he set up around the AI.

  9. 沉默王二75

    Claude Opus 5.5 提示词指南实践：删掉 think carefully、按任务调 effort

    作者根据 Claude 官方新出的提示词指南，整理出在 Claude Code 中使用 Opus 5.5 的几条做法。官方称 Opus 5.5 每次回复前都会自行决定思考多少，删掉提示词里的 think carefully 后回复更早且质量没有下降，作者用 grep 命令清理了 CLAUDE.md、rules 和 Skill 中的这类指令。

    Awaiting translation

9/28Mon
  2. Habr · Claude Code62

    按每小时 token 消耗给五家 API 网关排成本：Claude Opus 5.5 与 GPT-6 Sol 对比

    作者导出自己 49 个 Claude Code 会话、138 小时活跃工作的日志，统计出每小时平均消耗 1530 万缓存读取、86 万缓存写入、5.1 万输出和 560 普通输入 token，再按这套固定用量给 TeamoRouter、LiteAI、RouterAI、Polza AI、ProxyAPI 五家 API 网关算每小时花费。

    Awaiting translation

  3. Habr · Claude Code26

    Claude Code 如何帮我用手机测音和自建遥控器调好阁楼音响

    营销人 Anton Budon 用 Claude Code 配合三星 S23 上的 Android 声级计 App，测出阁楼听音位在 125 赫兹处有低频隆起，并发现功放 loudness 按钮会整体抬高 7 分贝低频、参考音变化会让曲线失真。随后他让 Claude Code 接入 Яндекс 音乐 API，做了一个局域网遥控器，并在其中加入针对该频段削减分贝的均衡器。

    Awaiting translation

9/27Sun
  1. AI-Driven Development · Родион Мостовой60

    用 BugSink 触发云端 Claude Code 自主修 bug，并补齐 vibe 项目的可观测性

    作者在实验用 BugSink 在异常发生时触发云端 Claude Code，让它读取生产环境只读数据和仓库、定位原因并提交修复，项目为实验性质、容错成本低。作者强调这套流程依赖透明的日志、追踪和遥测，而代码智能体默认做不好这些。

    Awaiting translation

9/26Sat
  4. Boris Cherny60

    Claude Tag 在 Slack 中现已支持个人连接器，可直接访问个人有权限的 Drive 文档、Salesforce 账号或数仓表，今天在 Teams 上线、下周面向 Enterprise 开放。

    Awaiting translation

    QuotedNoah Zweben@noahzweben

    Claude Tag in Slack can now use your personal connectors! You can now securely access that Drive doc, Salesforce account, or Warehouse table that you have personal access to right where the work happens. Avail. on Teams today and Enterprise next week https://claude.com/blog/claude-tag-now-supports-personal-connectors-in-channels

9/25Fri
  4. Sean Goedecke · Blog65

    为什么你应该在对话中多提问，包括对 AI 智能体

    作者主张在别人讲解方案时平均每三十秒问一个确认性问题，因为早期的小误解会层层放大，等到讲完再一起问就来不及了。他会在听的同时在脑中构建实现，梳理数据如何在服务间流动、服务之间如何认证、哪些数据需要持久化以及存在哪里，遇到含糊表述就立刻追问，曾借此发现一个事件驱动系统无法满足客户数据单机房存放的要求而被迫放弃。

    Awaiting translation

9/24Thu
  2. Hacker News · Prompt Injection78

    Can Open-Source Prompt Injection Detectors Stop Real AI Agent Attacks? Testing 629 AgentDojo Attacks in Practice

    The author tested 10 open-source prompt injection detectors against 629 AgentDojo injection attacks—each buried in real tool output—plus 97 benign samples.

    Why it matters: The author tested 10 open-source detectors against 629 real injection attacks, with full comparison data at both default thresholds and after calibration.

9/23Wed
  3. Tproger · Программирование80

    Anthropic Releases Flagship Model Claude Opus 5.5

    On September 22, Anthropic released its flagship model Claude Opus 5.5, aimed at developers and teams who want agents to handle multi-step tasks like coding and data analysis. The company says it delivers better performance and lower cost than Opus 5.

    Why it matters: Anthropic's published pricing and the default workload cost reduction help developers estimate the migration cost for long-running agent tasks.

  5. Lovable · Blog60

    Lovable Ships Opus 5.5: Faster Builds, Quality on Par with Opus 5

    Lovable has shipped Opus 5.5, which the company says matches Opus 5 in results while cutting the number of steps by one-third to one-half. On Lovable's internal benchmarks, Opus 5.5 ties Opus 5 on 0-to-1 builds and iterative code changes, and comes out 4% to 6% ahead on validation discipline; across all reasoning effort levels, steps per task drop by 26% to 57% and input tokens fall by 21% to 59%, with the differences significant at the 95% confidence level.

    Why it matters: Lovable shares official comparison data between Opus 5.5 and Opus 5 on step counts and tokens, so readers can judge the real change in build efficiency.

9/22Tue
  2. AI Coder · Telegram78

    Role-Based Model Routing in Claude Code: One Week of Practice and Hook Enforcement

    Drawing on his own Codex setup, the author built a role-based model routing system for Claude Code: the main thread acts as coordinator, explorer uses Haiku for code search only, worker uses Opus for TDD implementation, verifier uses Sonnet to run checks independently, senior uses high-tier Opus for money, data, and concurrency, and reviewer switches to a different model for semantic review.

    Why it matters: Based on a week of hands-on testing, the author shares the configuration, Hook enforcement, and cost trade-offs of multi-model division of labor in Claude Code, which you can adapt to your own multi-agent workflows.

9/21Mon
  2. Hacker News · Vibe Coding 讨论58

    用智能体六周做出 Pi Pocket：一次 Vibe Coding 实践复盘

    作者用智能体从零开发了 pi 的移动端前端 Pi Pocket，六周内达到 300 次提交，如今几乎不再逐行审查代码。他认为智能体适合边界清晰的小任务、重构和规划，但产出的代码普遍过度设计，规模比自己写大约 10-20%，样式和文档也偏模板化，且模型在主观判断上总顺着用户。作者目前只把这种方式用于个人项目，工作代码仍以 Claude 生成为主，自己写的不到 1%。

    Awaiting translation

9/19Sat
  1. Simon Willison · Coding Agents65

    Claude Code 2.1.277 起支持 AGENTS.md

    Claude Code 从 2.1.277 版本开始支持 AGENTS.md：当文件夹中没有 CLAUDE.md 时，Claude 会检查并使用 AGENTS.md。该支持基于 Claude Code mods 构建，这是其即将推出的定制 Claude Code harness 的方式，属于内置 mod，用户之后也可以自行构建自定义版本的项目指令。

    Awaiting translation