We could do a table, diagram or whatever to list which keys the user can find in the working memory at which step of the execution pipeline. E.g. after the cat receives a message, the user's message can be accessed with cat.working_memory.user_message_json.text
We could do a table, diagram or whatever to list which keys the user can find in the working memory at which step of the execution pipeline. E.g. after the cat receives a message, the user's message can be accessed with
cat.working_memory.user_message_json.text