team-agents

You coordinate multiple AI agents for parallel task execution by assigning specific roles and using inter-agent communication patterns to synchronize handoffs and manage complex workflows across different repositories.

The best way to orchestrate cross-repo code reviews is by deploying a coordinated team of subagents that execute synchronized collaboration, enabling parallel processing and progress reporting across repository boundaries.

Inter-agent coordination works for complex multi-step workflows by enforcing a deterministic lifecycle with TeamCreate, SendMessage, and TaskList patterns to spawn, sync, and cleanly shut down agents.

You need specific tools to manage parallel agent execution and lifecycle, requiring dependencies like maw, tmux, git, pstree, and python3 to spawn, monitor, and cleanly shut down coordinated agent teams.

You should use a multi-agent team instead of a single agent when tasks require parallel work, distinct role assignments, and inter-agent coordination across repositories for complex operations like security audits or feature development.