team-builder

To assemble an AI agent team for parallel execution, use an interactive menu to discover and select agent personas from available collections, then spawn them simultaneously to process a given task.

Yes, you can merge agents from both flat and domain-subdirectory collections into an ad-hoc team, capping the selection at 5 agents to maintain manageability and optimal workflow automation.

Parallel task dispatch works by spawning selected agents simultaneously, allowing them to process the described task independently and collecting their separate results for final synthesis with built-in error handling.

The maximum number of agents you can select for a task is capped at 5 agents, ensuring the ad-hoc team remains manageable while still supporting parallel execution and result synthesis.

Yes, the team builder supports interactive domain-based discovery, allowing users to navigate domain menus, compare available personas, and merge selections into a working team tailored to a specific task.

If an agent fails during parallel execution, the system includes error handling mechanisms to manage the failure and synthesize results from the remaining independent outputs of the spawned agent team.