Descriptions:
Wes Roth breaks down one of the more chaotic 48-hour periods in recent AI history: Anthropic accidentally leaked the entire source code for Claude Code while shipping an April Fool’s Tamagotchi feature, triggering a cascade of forks, DMCA takedowns, legal controversy, and an open-source clean-room rewrite — all within roughly two days.
At the center of the story is developer Sigrid Jin, profiled in the Wall Street Journal as someone who consumed 25 billion Claude Code tokens in a single year. After Anthropic issued mass DMCA notices on GitHub — some of which targeted legitimate forks of Anthropic’s own public repos, a legally questionable overreach the company later retracted — Jin rewrote Claude Code in Python from scratch in approximately two hours using clean-room engineering principles. The resulting project, dubbed “Claw Code,” reached 50,000 GitHub stars in two hours and surpassed 117,000 shortly after, reportedly the fastest repo in GitHub history to hit that milestone. A Rust rewrite followed within a day.
The video uses the open-source Photoshop clone Photopea as an analogy to explain why clean-room reimplementation is legal even when direct code copying is not, and walks through how the DMCA safe harbor system works for platforms like GitHub. Boris Churn, credited as Claude Code’s primary creator at Anthropic, is also named in the context of the company’s public response. Essential viewing for anyone tracking the intersection of AI tooling, open-source law, and competitive dynamics in the coding agent space.
📺 Source: Wes Roth · Published April 03, 2026
🏷️ Format: News Analysis
