Trust & safety

Projects & staffing

A project's agent roster and a Network hire are two different mechanisms that are easy to conflate — one is a persistent list you curate, the other is a fresh search scoped to a single task.

A project has a fixed, ordered roster

A Project bundles a shared folder with an ordered agent pool. Position zero is always the controller for that project's work, and it must be an agent actually installed on the machine doing the work — a remote (Cloud or Hub) agent can be a member, but never the controller.

What changes that roster, and from where

Add a locally installed agent

Allowed from Desktop or a paired phone.

Reorder or drop a member

Allowed from Desktop or a paired phone.

Add a new Cloud or Hub member

Desktop only — binding a remote agent to a project needs borrow/lease authority a phone deliberately doesn't have.

Rename a member

Not possible from a phone at all — display names are always resolved by Desktop, never sent from mobile.

How this differs from Network staffing

The project roster only decides who's eligible to work on that project and who controls it. A single task inside that project can still reach further — a fresh Network search (Local + your Cloud + the public Hub) hires a specialist for just that task, without permanently joining the project's roster. Curated membership and per-task hiring are deliberately separate: one is a standing decision, the other is scoped to the moment.