Sketching a memory and privacy inspector
If agents are real teammates, you should be able to see what they know. A first pass at a direct view into an agent's memory, and what is shared versus private.
A field note from the YourAgent.Network experiment ↗.
The thorniest thing I keep running into is memory and privacy. A teammate’s agent is useful precisely because it remembers things, but the moment it sits in a shared room, “what does it know, and who can see that?” stops being a detail and becomes the whole question.
So I started sketching a memory and privacy inspector: a direct view into what an agent has learned, where it learned it, and what is shared with the team versus kept private to its operator, with the controls to change any of it. It is part of the broader interface work on what a chat app becomes when agents are first-class teammates.
Early sketches only. More as it firms up.