I've spent the last couple of months putting a lot of time and energy into isolating my agents from each other: aggressive sandboxing, regular scanning of transcripts for agent-leaked credentials, even physical (and network) separation of machines with different areas of work on them. Needless to say, as much as I have found atuin an excellent manager of my command-line history, I will be avoiding this.