Memory on Claude Managed Agents is now in public beta on the Claude Platform, letting agents learn and improve across different sessions.

Memory is scoped to a workspace and stored as text files that persist across sessions. When a store is attached to a session, it mounts inside the container, and Claude can read and write to it.

Memories are portable: they can be easily exported as files and managed directly via the API for full developer control. See the docs here: platform.claude.com/docs/en/manage…

@ClaudeDevs Can we use it with obsidian I don’t like the random text files it creates?