v2.5

Teams can accumulate and persist knowledge over time, improving their responses and decisions across sessions and runs.

This brings organizational learning to multi-agent systems. Teams retain context about goals, constraints, policies, and past outcomes, leading to better performance and consistency as usage scales.

Details

• Teams now support LearningMachine for persistent knowledge retention
• Learned knowledge persists across sessions and is shared among team members
• Enables continuous improvement without manual re-prompting or context injection

Who this is for: Organizations running long-lived multi-agent teams that benefit from accumulated context, such as customer support, internal operations, or advisory systems.

View the Learning Machines docs.