# Product Log

## 2026-06-16

Privacy is the central future issue. FollowBrain should solve user-owned privacy first, then become the foundation for personal agents.

Canonical product framing shifted to **Mini Me**: a private personal agent that knows the user well enough to act in the real world. First proof should be narrow: "order my usual" for one food flow.

Key insight:

- Each person has a private second brain and personal agent.
- Agents can coordinate with each other without revealing raw private data.
- Example: "Schedule coffee with ___ when we are both available."
- User agent and other person's agent negotiate availability and place a meeting on calendars.
- They reveal minimum necessary information: compatible time slots, not full calendars.
- User chooses what data to share with big apps/companies.
- Big apps request scoped access; user grants, denies, expires, revokes, or monetizes access.

Product implication:

- Build private agent-to-agent coordination into mission.
- Treat app/company access as permissioned data licensing, not default extraction.
- Privacy is not a feature toggle. It is the foundation.