dispatching-parallel-agents

Debug multiple failing tests simultaneously by dispatching parallel AI agents to investigate independent failures without shared state. This skill coordinates isolated agents and aggregates their summaries for safe integration.

Yes, you can troubleshoot separate subsystems concurrently by dispatching parallel agents. The skill enforces task boundaries and ensures each agent handles a self-contained objective without sequential dependencies or shared state.

The best way to investigate concurrent bugs is dispatching parallel agents to handle independent problem areas. It identifies distinct domains, assigns focused tasks, and merges agent summaries safely for integration.

No, you do not need to manually group failures. The skill performs independent domain identification automatically, grouping unrelated failures by distinct problem areas before assigning focused agent tasks.

You should not use parallel agents for dependent issues because the skill enforces strict task boundaries without shared state. It is designed specifically for independent problems requiring isolated investigation and summary aggregation.

To start dispatching agents, simply ask the skill to dispatch a separate agent for each independent failure you list. The skill coordinates the concurrent execution and aggregates their summaries.