What problem does it solve?
This Skill empowers the agent to act as a Swarm Leader, enabling it to delegate tasks to other AI agents (Workers) within a cluster, manage their execution, and consolidate results, thereby overcoming the limitations of a single agent's capacity.
Core Features & Use Cases
- Task Delegation: Distribute complex tasks to specialized or available Worker agents.
- Concurrency Management: Utilize
dispatch_batch_tasks for parallel execution of independent tasks, significantly speeding up workflows.
- Stateful Collaboration: Maintain context across multiple turns with specific Workers using
sub_session_id for iterative development or problem-solving.
- Information Gathering: Employ
sync_task_context to query the status and progress of tasks across the swarm.
- Collaborative Discussion: Facilitate multi-agent brainstorming and decision-making with
hold_meeting.
- Robust Execution: Features
deep_think for complex, verifiable problem-solving and fallback mechanisms for agent failure.
- Use Case: When asked to "research 5 competitors and summarize their Q3 strategies," the agent can use
dispatch_batch_tasks to assign each competitor research to a different Worker, then use sync_task_context to gather results, and finally synthesize a comparative report.
Quick Start
Use the agent_team skill to dispatch the task 'Analyze the provided user data for anomalies' to an available worker.
