内容来源：https://www.theverge.com/ai-artificial-intelligence/904776/anthropic-claude-source-code-leak

内容总结：

近日，人工智能公司Anthropic在其编程工具Claude Code的2.1.88版本更新中意外泄露了内部源代码，涉及代码量超过51.2万行。此次泄露并非由外部攻击导致，而是因发布打包过程中的人为失误所致。

据科技媒体Ars Technica和VentureBeat报道，泄露的代码中包含了未发布的功能设计、AI指令集以及系统“记忆”架构等核心信息。有用户在代码中发现了一个类似“电子宠物”的交互功能，该设计可伴随用户编程过程产生反应；同时代码中还提及名为“KAIROS”的常驻后台代理模块。此外，代码注释中甚至出现了开发人员对部分设计复杂性的质疑。

尽管Anthropic在发现问题后迅速修复，但相关代码已被用户上传至GitHub平台，并衍生出超过5万个复刻版本。公司发言人克里斯托弗·纳尔蒂强调，此次事件未涉及用户敏感数据或凭证泄露，并已启动措施防止类似情况再次发生。

行业分析师指出，此类泄露虽可能带来技术细节外泄风险，但也将促使企业加强内部流程管控。Claude Code自2025年2月发布以来，因其具备代理执行任务的能力而备受关注，此次事件或将成为其完善运营体系的重要节点。

中文翻译：

Anthropic发布Claude Code的2.1.88版本更新后，用户很快发现其中包含一个带有源码映射文件的软件包，内含其TypeScript代码库。X平台上有用户率先指出此次泄露事件，并上传了包含代码的文件。据Ars Technica和VentureBeat早前报道，泄露数据包含超过51.2万行代码，揭示了这款AI编程工具的内部运作机制。

Claude代码泄露事件曝光了电子宠物式"伙伴"与常驻后台智能体

超过51.2万行的泄露代码似乎展示了未发布功能、Claude指令集等内部信息。

深入分析代码的用户声称发现了即将上线的新功能、Anthropic为AI助手设定的指令体系，以及其"记忆"架构的运作逻辑。Reddit论坛帖子显示，用户发现的亮点包括：类似拓麻歌子的虚拟宠物"驻留在输入框旁并对编程行为作出反应"，以及可能实现常驻后台智能体的"KAIROS"功能。用户还发现了Anthropic程序员的一条注释，其中坦言"此处的记忆化处理大幅增加了复杂度，且不确定是否真正提升了性能"。

尽管Anthropic随后修复了该问题，但用户仍将代码复制到GitHub仓库，该仓库目前已产生超过5万个分叉版本。Anthropic于2025年2月推出Claude Code，在新增可代用户执行任务的智能体功能后，该工具获得了更广泛的市场关注。

Anthropic发言人克里斯托弗·纳尔蒂通过邮件向The Verge声明："今日稍早的Claude Code版本意外包含了部分内部源代码，但未涉及或泄露任何客户敏感数据与凭证。这属于人为失误导致的版本打包问题，并非安全漏洞。我们正在推行改进措施防止此类事件再次发生。"

高德纳咨询公司AI分析师阿伦·钱德拉塞卡兰向The Verge表示，虽然Claude代码泄露可能"为恶意行为者提供绕过安全防护的潜在途径"，但其长期影响或许仅限于"促使Anthropic加大流程与工具投入，提升运营成熟度"。

英文来源：

After Anthropic released Claude Code’s 2.1.88 update, users quickly discovered that it contained a package with a source map file containing its TypeScript codebase, with one person on X calling attention to the leak and posting a file containing the code. The leaked data reportedly contains more than 512,000 lines of code and provides a look into the inner workings of the AI-powered coding tool, as reported earlier by Ars Technica and VentureBeat.
Claude Code leak exposes a Tamagotchi-style ‘pet’ and an always-on agent
The more than 512,000 lines of leaked code appear to show unreleased features, instructions for Claude, and more.
The more than 512,000 lines of leaked code appear to show unreleased features, instructions for Claude, and more.
Users who have dug into the code claim to have uncovered upcoming features, Anthropic’s instructions for the AI bot, and insight into its ”memory” architecture. Some things spotted by users include a Tamagotchi-like pet that “sits beside your input box and reacts to your coding,” according to a post on Reddit, along with a “KAIROS” feature that could enable an always-on background agent. Users also found a comment from one of Anthropic’s coders, who admits at one point that the “memoization here increases complexity by a lot, and im not sure it really improves performance.”
Though Anthropic later fixed the issue, that didn’t stop users from copying the code to a repository on GitHub, which has since amassed more than 50,000 forks (or copies of the repository). Anthropic launched Claude Code in February of 2025, and the tool picked up more steam after adding agentic capabilities that perform tasks on a user’s behalf.
“Earlier today, a Claude Code release included some internal source code. No sensitive customer data or credentials were involved or exposed,” Anthropic spokesperson Christopher Nulty says in an emailed statement to The Verge. “This was a release packaging issue caused by human error, not a security breach. We’re rolling out measures to prevent this from happening again.”
Arun Chandrasekaran, an AI analyst at Gartner, tells The Verge that while the Claude Code leak poses “risks such as providing bad actors with possible outlets to bypass guardrails,” its long-term impact could be limited to serving as a “call for action for Anthropic to invest more in processes and tools for better operational maturity.”