What does it mean that shared state between agents?

When agents need to share information define clearly how that information flows.

Example: You run a morning system with three agents: a planning agent that sets your daily priorities, a calendar agent that blocks time for deep work, and an energy-tracking agent that logs how you felt after each work block. Each agent does its job. But none of them can see what the others produced. The planning agent sets priorities without knowing your energy patterns. The calendar agent blocks time without knowing today's priorities. The energy tracker logs data that never reaches the planner. You have three competent agents operating on three private realities. Compare this to a version where all three agents read from and write to a shared daily state — a single document containing today's priorities, the current calendar, and yesterday's energy data. Now the planner accounts for energy patterns. The calendar respects priority order. The energy tracker's output feeds tomorrow's planning. Same agents, same capabilities. The difference is shared state.

Try this: Identify two or three agents — cognitive routines, tools, or processes — that you run regularly and that should inform each other but currently do not. Write down what each agent produces as output and what each agent would need as input to perform better. Then design a shared state artifact: a single document, dashboard, or note that all agents read from and write to. Use it for one week. At the end of the week, note which agents made better decisions because they had access to information they previously lacked.