Anthropic's Claude Code source leaked, revealing secrets

Anthropic's valuable Claude Code AI product has suffered a significant source code leak, exposing its internal workings and proprietary architecture. The leak, which occurred via a debugging file on the npm registry, provides competitors with a detailed blueprint for building advanced AI agents, particularly concerning its novel memory management system designed to combat 'context entropy.' This incident represents a substantial loss of intellectual property for Anthropic, potentially impacting its competitive edge in the rapidly evolving AI market and raising security concerns for users.