Anthropic Leak Reveals 4 Hidden Claude Features Redefining AI Future

• A recent leak exposed unreleased features for Anthropic's AI coding assistant, Claude, due to a "release packaging issue caused by human error."
• The leak revealed four secret features: BUDDY (gamified AI companion), KAIROS (always-on autonomous Claude with a "dreaming cycle"), ULTRAPLAN (cloud-based deep work planning), and COORDINATOR MODE (multi-agent collaboration).
• BUDDY is a Tamagotchi-style AI pet with 18 species and evolving personalities, targeting a May 2026 launch.
• KAIROS enables Claude to persist across sessions, monitor activity, and proactively take actions, functioning as a continuous-learning background agent.
• COORDINATOR MODE allows one Claude instance to spawn multiple sub-agents, creating an orchestrated multi-agent system for distributed AI workflows.