Leak of Claude Code Source Code via npm

<p>Today, Chaofan Shou discovered that the npm package @anthropic-ai/claude-code version 2.1.88 contained a 59.8 MB source map file - essentially the complete source code of the project. Within a few hours, 512,000 lines of TypeScript code spread across GitHub (41,500+ forks).</p>
<p>Anthropic confirmed: "This is a release packaging error caused by human factors, not a hack."</p>
<p>What was found inside:</p>
<ul>
<li>44 feature flags with hidden features that are already written but not enabled</li>
<li>KAIROS - an autonomous agent mode that runs 24/7 in the background, subscribes to GitHub webhooks, and can "sleep" and resume work by itself</li>
<li>BUDDY - a built-in tamagotchi with 18 types of creatures (duck, dragon, axolotl, capybara), a rarity system, and RPG stats like CHAOS, SNARK, and DEBUGGING</li>
<li>Voice mode - full voice control via push-to-talk</li>
<li>Browser control via Playwright</li>
<li>Multi-agent mode - one Claude launches several working Claudes with limited toolsets</li>
</ul>
<p>Unexpectedly, the code also contained an "undercover mode" that removes all traces of AI-written code when Anthropic employees commit to open source projects. They also discovered fake tools - decoy tools for data poisoning if someone records API traffic to train a competing model. And regex for detecting user frustration through profanity - in a company that makes LLM.</p>
<p>Technically - 1,900 TypeScript files, runs on Bun (not Node.js), terminal UI on React + Ink, 40 built-in tools, 50 slash commands.</p>
<p><a href="https://x.com/fried_rice/status/2038894956459290963">https://x.com/fried_rice/status/2038894956459290963</a></p>;
<p><a href="https://github.com/Kuberwastaken/claude-code">https://github.com/Kuberwastaken/claude-code</a></p>;
<p>#anthropic #claude #claude_code #ai #security #leak</p>
<p>—————————<br>
Мысли Рвачева<br>
—————————</p>