Anthropic accidentally published 512,000 lines of proprietary source code from its flagship AI coding tool Claude Code through a debug file bundled into an npm update on Mar. 31, exposing the full architecture of a product that generates an estimated $2.5 billion in annualized revenue and raising fresh operational questions for a company valued at $350 billion.
Claude Code Leak Details
Security researcher Chaofan Shou discovered the exposed source map file in Claude Code version 2.1.88 and posted a download link on X. The codebase spread across GitHub within hours, accumulating tens of thousands of forks before Anthropic's DMCA takedowns landed.
The incident marked the second accidental disclosure in five days.
A separate CMS misconfiguration on Mar. 26 had already exposed roughly 3,000 internal files, including details on the unreleased "Mythos" model.
Anthropic confirmed the leak to multiple outlets, calling it "a release packaging issue caused by human error, not a security breach." Enterprise clients, who account for 80% of Claude Code's revenue, now face a tool whose internal security logic sits on the open internet.
Also Read: XRP Binance Outflows Near 6,000 Daily Transactions
Sigrid Jin's Rewrite
Korean developer Sigrid Jin, profiled by The Wall Street Journal for consuming 25 billion Claude Code tokens last year, completed a clean-room Python rewrite before sunrise. His repository, claw-code, hit 50,000 GitHub stars within two hours.
The leaked files revealed an internal feature called "Undercover Mode," built to prevent Claude from exposing Anthropic's internal details in public repositories.
The code also contained 44 feature flags, an unreleased background daemon called KAIROS, and internal model codenames — including "Capybara" for a Claude 4.6 variant.
Two accidental disclosures in a single week raise operational questions for a company reportedly considering an IPO in Q4 2026.
AI Token Fallout
The broader AI cryptocurrency sector had already taken a hit from the Mar. 26 Mythos leak, which sent cybersecurity stocks lower and pushed Bitcoin (BTC) back to $66,000. The AI crypto sector, currently valued at roughly $17.7 billion, shed 1.5% that week alone.
Repeated security lapses at a leading centralized AI lab could cut both ways for AI-linked tokens such as Bittensor (TAO), Render (RNDR), and the Artificial Superintelligence Alliance (FET).
On one hand, the incidents strengthen the narrative that decentralized AI infrastructure offers a more resilient alternative to corporate labs. On the other, a loss of confidence in the broader AI sector's operational maturity could drag token prices down alongside traditional tech equities — as it did when the Mythos leak first sent shockwaves through the market.
Read Next: Midnight Mainnet Debuts On Cardano With 9 Partners, Including Google Cloud