moai-alfred-agent-guide

Multi-agent orchestration uses decision trees to route each task to the appropriate sub-agent based on task type and complexity. This Skill provides decision-tree logic for selecting among 19 sub-agents, ensuring each task reaches the agent best suited to handle it—whether for feature development, bug triage, code exploration, or documentation work.

Model selection depends on task complexity and latency requirements. Haiku handles simpler, faster tasks efficiently, while Sonnet tackles reasoning-heavy work. This Skill specifies model-choice criteria within Alfred workflows so agents automatically select the right model for each sub-task, balancing speed and capability.

Clear responsibility boundaries prevent duplicate work and ensure efficient collaboration. This Skill documents functional requirements and collaboration patterns for 19 sub-agents, defining what each agent owns and how agents coordinate when tasks span multiple domains.

Decision trees systematically narrow down agent choices by evaluating task attributes—scope, urgency, domain—against explicit rules. This Skill applies decision-tree routing across feature development, bug triage, code exploration, and documentation synchronization, eliminating guesswork and ensuring consistent dispatch logic.

While possible, predefined decision rules ensure reliability and consistency at scale. This Skill provides explicit decision-tree specifications and responsibility boundaries for orchestrating 19 sub-agents, reducing errors and operational overhead compared to ad-hoc routing.