2 hours, 50,000 stars; in one day, surpassing 100,000: Claude Code leak triggers the fastest project in GitHub history, rewritten from scratch to evade DMCA

According to 1M AI News monitoring, within less than 24 hours after the Claw Code source code leak, a project called Claw Code amassed more than 100,000 stars on GitHub at a record pace—according to the project’s own description, it is the fastest repository in GitHub history to reach this milestone. As of the time of publication, the project has received 124,000 stars and 102,000 forks. Unlike the thousands of repositories that simply rehost leaked source code, Claw Code’s approach is to only look at the architecture and not copy the code, implementing it fully from scratch—engineered and legally referred to as a “clean-room implementation.”

Project author Sigrid Jin described this experience in the README: at 4:00 a.m. on March 31, the leak hit and triggered an explosion of notifications on his phone. In South Korea, his girlfriend worried that “there could be legal risk just because the computer has these codes on it.” He made the kind of decision an engineer under pressure would make: he sat down, rewrote the core functionality from scratch in Python, and pushed it to GitHub before daybreak. The entire process was assisted by an AI coding workflow tool, oh-my-codex (OmX), developed by Yeachan Heo.

This is the key reason Claw Code was not taken down by Anthropic under the DMCA. Anthropic sent copyright protection notices to 8,100 repositories that directly mirrored leaked source code, but there isn’t a single line of original TypeScript in Claw Code’s codebase. Instead, it was rewritten in Python first, then shifted to Rust; currently, the codebase is 92.9% Rust. The project includes 7 Rust crates covering an API client, runtime state management, MCP orchestration, a tool execution framework, a plugin system, a command system, and an interactive CLI.

Sigrid Jin is not an unknown developer. In a March 21 report by The Wall Street Journal titled “The Trillion Dollar Race to Automate Our Entire Lives,” he was portrayed as one of the globe’s most active heavy users of Claude Code, saying that last year he used 25 billion Claude Code tokens by himself and that he even flew to San Francisco to attend Claude Code’s one-year anniversary gathering.

Earlier, the fastest GitHub project to reach 100,000 stars reportedly was OpenClaw, which took weeks. Claw Code compressed that timeline to within a day, but GitHub does not publish official growth rankings; this record comes from the project’s own account and citations by multiple technology media outlets. It’s also worth noting that the project’s fork count (102,000) to star count (124,000) ratio exceeds 80%, far higher than the 10%-20% level typical for normal open-source projects. These stars are more often driven by actions sparked by the leak event rather than actual product usage of the project itself.