Claude Code leak exposes a Tamagotchi-style ‘pet’ and an always-on agent

The Verge / 4/1/2026

📰 NewsDeveloper Stack & InfrastructureSignals & Early TrendsTools & Practical Usage

Key Points

  • After Anthropic shipped a Claude Code update (2.1.88), a source map file included in a distributed package allegedly exposed the tool’s TypeScript codebase.
  • The reported leak includes over 512,000 lines of code, giving researchers and users a window into Claude Code’s implementation details.
  • Community findings reportedly point to upcoming features, detailed instructions for the agent, and insight into its “memory” architecture.
  • Multiple outlets (including Ars Technica and VentureBeat) and social media posts helped surface and circulate what was discovered, turning the update into an immediate security and product intelligence story.
The Claude logo on a beige background.

After Anthropic released Claude Code's 2.1.88 update, users quickly discovered that it contained a package with a source map file containing its TypeScript codebase, with one person on X calling attention to the leak and posting a file containing the code. The leaked data reportedly contains more than 512,000 lines of code and provides a look into the inner workings of the AI-powered coding tool, as reported earlier by Ars Technica and VentureBeat.

Users who have dug into the code claim to have uncovered upcoming features, Anthropic's instructions for the AI bot, and insight into its "memory" architecture. Some things spotted by users inclu …

Read the full story at The Verge.