working-with-subagents

Task delegation to subagents for parallel execution is coordinated by assigning explicit agent IDs and prompts to stateful workers, which route tasks based on type and resource constraints while preserving shared memory across calls.

Stateful subagent orchestration coordinates persistent workers like scout, coder, and memory agents to execute delegated tasks without losing context. You need it for sustained sessions requiring memory-driven reasoning and parallel work across multiple specialized agents.

Preserving context and memory across multiple AI agent calls requires stateful subagents that maintain memory blocks between executions. The system routes decisions based on task type, ensuring shared memory persists throughout sustained sessions.

Parallel task execution with persistent scout, coder, and memory agents is supported by coordinating stateful subagents. You must provide explicit agent IDs and prompts, allowing the system to deploy workers and monitor shared memory results.

Memory-driven reasoning in task delegation requires explicit agent IDs and prompts for routing decisions. Limitations include the need to manage resource constraints and task type routing manually, as the system depends on sustained session state to preserve context.