▸ AI/LLM · Axios
Anthropic Leaks Claude Code Source: 500K Lines Exposed
Anthropic accidentally published its Claude Code source code via a misconfigured npm package, exposing 512,000 lines of TypeScript across 1,906 files. The leak revealed unreleased features including a 'KAIROS' persistent agent mode and internal model codenames. This marks Anthropic's second major data exposure in under a week.
The bottom line: Competitors now have a detailed blueprint for building production-grade AI coding agents, and the incident raises questions about operational security at the 'safety-first' AI lab.