Anthropic's "Code with Claude" event revealed several major updates to its AI platform. One key feature is scheduled AI routines, which allow users to automate recurring workflows based on schedules or webhooks. This means developers can now use Claude Code to integrate AI into their products and services in a more seamless way.
Another significant update is the introduction of outcome-based agents, which enable rubric-based agent grading. This allows developers to evaluate an AI's performance based on specific criteria, making it easier to build and train effective agents. Anthropic has also introduced multi-agent orchestration, enabling specialized AI teams with different roles and tools.
Additionally, the platform now features a new memory system called "Dreams," which is designed for long-term agent behavior. This update allows developers to create more sophisticated and realistic AI interactions. With these updates, Anthropic's platform is becoming increasingly competitive in the market, particularly with its rival OpenAI's Codex platform.
The usage limits on Claude Code have been increased, allowing developers to build more complex products without running into limitations.
