agent amnesia and the purpose of memory files in .openclaw/agents and .openclaw/workspace /agents
I have gone from a single agent setup to a council of agents. I am looking through my openclaw files, and I am seeing what seems like a duplicatioin of files between a folder in the .openclaw root called agents and another folder inside workspace/agents I am seeing two sets of files, with MEMORY.md files showing up in both places and memory folders. Can you tell me if this is a mistake, and I should aggregate these into a single location, or if they have different purposes, and what they are. I have issues even within a single agent within sessions, where one session knows thing that the other session does not know about, like the access protocol to a local database. I would like for the agents to have more consistent memories, and I think this may be one reason that they are so inconsistent. Can you also review the most basic filestructure that every agent is born with and where those files live?
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
134,827Members
Resources
Similar Threads
Was this page helpful?
