Updated:12:47 AM CEST Jun,04

NEWS / RSS
CHEATS

REVIEWS

ARTICLES

HARDWARE

DEMOS

FORUMS

LINKS (new)

66 lottery login

91 club

okwin

bdg game

55 club

CONTACT

Please e-mail us if you have news.

Anthropic's Epic Blunder: Claude Code Source Leaked, Then Rewritten Overnight in Record Time - tech

(hx) 12:18 AM CEST - Apr,01 2026

Anthropic accidentally exposed roughly 512,000 lines of proprietary TypeScript source code for its Claude Code AI coding tool when a debug source map file was bundled into an npm package update early Tuesday morning.

Researcher Chaofan Shou quickly spotted the 59.8 MB file and shared a download link on X, where the thread drew millions of views, allowing the code to be downloaded and mirrored globally before Anthropic could respond.

The company pulled the package and issued DMCA takedowns, describing the incident as a "release packaging issue caused by human error" with no customer data exposed. Korean developer Sigrid Jin, known for using 25 billion Claude tokens last year, responded by clean-room rewriting the core in Python as "claw-code" before sunrise, creating a new work immune to DMCA claims; the repo surged to over 30,000 GitHub stars rapidly, with a Rust port following and amassing tens of thousands more.

The original code has also been mirrored on decentralized platforms, ensuring it remains publicly accessible despite takedown efforts and highlighting the challenges of controlling AI tooling in the open internet era.

Anthropic accidentally leaked their entire source code yesterday. What happened next is one of the most insane stories in tech history.

> Anthropic pushed a software update for Claude Code at 4AM.

> A debugging file was accidentally bundled inside it.

> That file contained… pic.twitter.com/4C4q3b6sOb
— Jeremy (@Jeremybtc) March 31, 2026

last 10 comments:

NakedFaerie	(11:06 AM CEST - Apr,01 2026 )
So what is all this? Anthropic Claude? Just another AI system? ALL AI is bad so dont replicate it, just scrap it.

Csimbi

(09:57 AM CEST - Apr,02 2026 )

I guess now the americans don't feel so bad about showing the finger to the government. The chinese and russians would hack their military AI to pieces with this.
But yes, it's just another AI - nothing to see here.
Are you guys aware of the billions invested by superscalers and only a fraction of the capacity being used? I think investors are already considering pulling the plug and when the first big investor pulls the trigger, the AI bubble will burst. It's about time we return to normality, if you ask me.

All comments

Add your comment (free registration required)