skill-parallel-orchestration

Parallel orchestration works by splitting a main objective into discrete sub-task files and asynchronously spawning independent sub-agents. It manages background process initiation and synchronizes agent progress through session state monitoring.

Task decomposition in multi-agent orchestration is the process of breaking down a complex main objective into discrete, manageable sub-task files. These atomic files are then assigned to spawned agents for concurrent execution and background processing.

You synchronize progress across multiple background agents by monitoring a shared session state file. The orchestration system reads this update_state.py dependency to track the completion status of each concurrently running sub-agent.

Yes, you can use parallel orchestration to build frontend and backend components concurrently by spawning separate agents for each sub-task. For example, one agent handles the frontend login page while another develops the backend simultaneously.

No, you do not need to manually manage background processes when spawning multiple agents. The orchestration skill initiates background processes via a mock runner script and handles the asynchronous execution without blocking your main process.

You should not use parallel task execution when sub-tasks are interdependent and require sequential processing. Since this orchestration approach splits work into independent atomic files, it is ineffective for work units that share state or block each other during development.