MoAI ADK - Multi-agent Orchestration Interface

Purpose

Orchestrate multiple AI agents for complex tasks using the MoAI framework.

Core Concepts

Agent Roles

• Planner: Breaks down complex tasks
• Executor: Performs specific actions
• Reviewer: Validates outputs
• Integrator: Combines results

Workflow Orchestration

1. Task decomposition
2. Agent assignment
3. Parallel execution
4. Result aggregation
5. Quality verification

MoAI Tool Integration

Tools Available

• Task dispatch
• Context sharing
• Result aggregation
• Conflict resolution

Usage Patterns

Pattern 1: Sequential Pipeline

Input → Agent A → Agent B → Agent C → Output

Pattern 2: Parallel Processing

        ┌→ Agent A →┐
Input → ├→ Agent B →┼→ Integrator → Output
        └→ Agent C →┘

Pattern 3: Review Loop

Input → Executor → Reviewer → (Approved → Output)
                          └→ (Rejected → Executor)

Best Practices

1. Clear Interfaces: Define inputs/outputs for each agent
2. Context Management: Share relevant context between agents
3. Error Handling: Plan for agent failures
4. Progress Tracking: Monitor multi-agent workflows
5. Result Verification: Validate final outputs

OpenCode Integration

Works seamlessly with OpenCode's subagent system:

• Use subagent tool for agent dispatch
• Leverage skill tool for capability loading
• Monitor with todo tool for task tracking

Usage

Activate for:

• Complex feature development
• Multi-step refactoring
• Cross-domain tasks
• Quality assurance workflows