some good discussions and experiments around stateful agents in the replies, but seems like we’re not quite there yet, as in we’re starting to track memory and traces, but not quite agent capability as part of that state
Technical assessment of stateful AI agent capabilities and memory
By
–