summon

Parallel tasks run concurrently by spawning multiple subagents simultaneously, with each agent seeded using the full skill registry to execute independent workstreams like code generation or testing.

Multi-agent orchestration coordinates parallel subagents to execute independent tasks simultaneously, applying deterministic task dispatch and per-agent context sharing to ensure standardized completion reporting.

Concurrent data processing executes by spawning parallel subagents designed for distributed testing, ensuring each agent receives shared context and reports standardized completion results.

Parallel subagents support multi-component feature development by spawning concurrent agents that handle independent tasks simultaneously while maintaining deterministic task dispatch and consistent capabilities.

Parallel subagent orchestration is ideal when you have several independent tasks that can run simultaneously, accelerating complex workflows compared to sequential execution for parallel-safe projects.

Spawning parallel subagents requires parallel-safe workstreams to function correctly; tasks with interdependencies or shared mutable state are not suitable for this deterministic task dispatch approach.