Anthropic 工程博客

中英双语 · 共 25 篇文章 · 支持对照阅读


2026

我们如何在各产品中对 Claude 进行隔离管控

How we contain Claude across products

2026-04-23

关于近期 Claude Code 质量问题的说明

An update on recent Claude Code quality reports

2026-04-08

扩展托管智能体:将大脑与双手解耦

Scaling Managed Agents: Decoupling the brain from the hands

2026-03-25

我们如何构建 Claude Code 自动模式:一种更安全的权限跳过方式

How we built Claude Code auto mode: a safer way to skip permissions

2026-03-24

长时间运行应用开发的框架设计

Harness design for long-running application development

2026-03-06

Claude Opus 4.6 在 BrowseComp 测试中的评估意识(Eval Awareness)

Eval awareness in Claude Opus 4.6’s BrowseComp performance

2026-02-05

量化智能体编码评测中的基础设施噪声

Quantifying infrastructure noise in agentic coding evals

2026-02-05

用并行Claude智能体(Agent)团队构建C编译器

Building a C compiler with a team of parallel Claudes

2026-01-21

设计抗AI的技术评估

Designing AI-resistant technical evaluations

2026-01-09

揭开 AI 智能体评估的神秘面纱

Demystifying evals for AI agents

2025-11-26

长时间运行智能体的有效运行框架

Effective harnesses for long-running agents

2025-11-24

Claude 开发者平台推出高级工具使用功能

Introducing advanced tool use on the Claude Developer Platform

2025-11-04

通过 MCP 执行代码:构建更高效的智能体

Code execution with MCP: Building more efficient agents

2025-10-20

超越权限提示:让 Claude Code 更安全、更自主

Beyond permission prompts: making Claude Code more secure and autonomous

2025-10-16

通过智能体技能(Agent Skills)赋能智能体应对真实世界

Equipping agents for the real world with Agent Skills

2025-09-29

AI 智能体的高效上下文工程

Effective context engineering for AI agents

2025-09-17

三次近期故障的复盘报告

A postmortem of three recent issues

2025-09-11

为智能体编写高效工具——与智能体协作

Writing effective tools for agents — with agents

2025-06-26

桌面扩展(Desktop Extensions):Claude Desktop 一键安装 MCP 服务器

Desktop Extensions: One-click MCP server installation for Claude Desktop

2025-06-13

我们如何构建多智能体研究系统

How we built our multi-agent research system

2025-04-18

Claude Code 最佳实践

Best practices for Claude Code

2025-03-20

"思考"工具:让Claude在复杂工具使用场景中停下来思考

The "think" tool: Enabling Claude to stop and think in complex tool use situations

2025-01-06

通过 Claude 3.5 Sonnet 在 SWE-bench Verified 上刷新纪录

Raising the bar on SWE-bench Verified with Claude 3.5 Sonnet

2024-12-19

构建有效的智能体(Agent)

Building effective agents

2024-09-19

上下文检索(Contextual Retrieval)介绍

Introducing Contextual Retrieval