Organizing agents in projects

Projects are where ACP agents become operational teams instead of isolated prompt profiles. Organizing agents in projects means deciding who owns what work, what computers they run on, and how they collaborate across tickets and threads.

Structure agent work at the project level so execution stays coordinated and reusable.

Assign agents to real work

ACP projects let tickets, threads, and mission control output be associated with specific agents or teams. That turns role definitions into execution behavior instead of leaving them as abstract configuration.

Assign the right agent to the right class of ticket.

Use teams when work should be orchestrated across roles.

Keep ownership visible at the project layer.

Align agents with computers and skills

A good project setup does not stop at assigning an agent name. The strongest ACP configurations also decide which computers and skills those agents should use so execution is repeatable and role-appropriate.

Pair agents with computers configured for their work.

Use skills to encode repeatable operational capability.

Keep the execution surface aligned with the role definition.

Coordinate project-scale execution

When projects hold multiple agents, teams, resources, and schedules, they become the control plane for operational collaboration. ACP uses projects to keep this coordination visible instead of hiding it across disconnected tools.

Use projects to coordinate tickets, agents, resources, and automation together.

Track execution history back to the agents responsible for it.

Turn project structure into a repeatable operating model for teams.