Why we build mini apps, not one big system
The goal is one agent that helps across all your work. We get there one small, checkable mini app at a time.
Aim big. Build in pieces small enough to prove.
Big goals need small, checkable steps
The long-term goal is what we call an omni-agent: one consistent relationship with your agents across what you know, the tools you use, the decisions you make, your work, your devices and other agents. If you build that as one huge product, you can’t tell which part is helping and which part is hiding a problem. So we measure progress in small, focused mini apps. Each one starts with a problem we kept running into while building, running, teaching or installing agents. Each one has to make that problem clearly better on its own.
Each mini app has one job
Task Board holds the work. Knowledge holds documents and where they came from. Rules & Approvals holds the rules. Operator Inbox holds the decisions a person has to make. The names matter less than the habit: no mini app quietly takes over another one’s job. Clear jobs make clear hand-offs between mini apps. They also let a client start with the one thing they need, check that it works, and add the next mini app only when the work calls for it.
Mini apps that work together without locking you in
Working on its own doesn’t mean working alone. Operator Inbox is useful by itself and stronger next to Rules & Approvals. Knowledge stands alone and gets richer next to Scribe, which reads chats and messages into it. Remote Setup can install one mini app or a whole client kit. The collection is a set of useful parts that are meant to fit together, not a bundle you have to buy whole. That is the practical route from today’s problem to tomorrow’s all-round agent.