Claude Code leak exposes a Tamagotchi-style ‘pet’ and an always-on agent
The Verge
A packaging error in a Claude Code update leaked internal source code, revealing experimental features like a virtual pet and an always-on agent.Summary
A recent update to Anthropic’s Claude Code tool inadvertently leaked over 512,000 lines of source code due to a packaging error. Users analyzing the leaked files discovered references to experimental features, including a Tamagotchi-style virtual pet and a background agent codenamed "KAIROS." Anthropic confirmed the incident was caused by human error rather than a security breach and stated that no sensitive customer data was exposed. While analysts suggest the leak could potentially help bad actors identify ways to bypass safety guardrails, Anthropic has since patched the issue and is implementing new measures to prevent future occurrences.
(Source：The Verge)