We argue for building agent-ready systems that allow AI agents to operate safely and efficiently in live environments with messy state, parallel users, and genuine service integrations — while maintaining visibility into side effects and enforcing safeguards.
External link: DAPLab Publications.